Split (1)

train · 838k rows

modelId stringlengths 5 139	author stringlengths 2 42	last_modified timestamp[us, tz=UTC]date 2020-02-15 11:33:14 2025-06-23 12:29:03	downloads int64 0 223M	likes int64 0 11.7k	library_name stringclasses 492 values	tags sequencelengths 1 4.05k	pipeline_tag stringclasses 54 values	createdAt timestamp[us, tz=UTC]date 2022-03-02 23:29:04 2025-06-23 12:24:08	card stringlengths 11 1.01M
Swadine/cartpole_policy_gradient	Swadine	2024-01-13T03:22:19Z	0	0	null	[ "CartPole-v1", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]	reinforcement-learning	2024-01-13T03:22:09Z	--- tags: - CartPole-v1 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: cartpole_policy_gradient results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: CartPole-v1 type: CartPole-v1 metrics: - type: mean_reward value: 474.00 +/- 78.00 name: mean_reward verified: false --- # Reinforce Agent playing CartPole-v1 This is a trained model of a Reinforce agent playing CartPole-v1 . To learn to use this model and train yours check Unit 4 of the Deep Reinforcement Learning Course: https://huggingface.co/deep-rl-course/unit4/introduction
ntc-ai/SDXL-LoRA-slider.waving-hello	ntc-ai	2024-01-13T03:21:41Z	100	1	diffusers	[ "diffusers", "text-to-image", "stable-diffusion-xl", "lora", "template:sd-lora", "template:sdxl-lora", "sdxl-sliders", "ntcai.xyz-sliders", "concept", "en", "base_model:stabilityai/stable-diffusion-xl-base-1.0", "base_model:adapter:stabilityai/stable-diffusion-xl-base-1.0", "license:mit", "region:us" ]	text-to-image	2024-01-13T03:21:38Z	--- language: - en thumbnail: "images/evaluate/waving hello.../waving hello_17_3.0.png" widget: - text: waving hello output: url: images/waving hello_17_3.0.png - text: waving hello output: url: images/waving hello_19_3.0.png - text: waving hello output: url: images/waving hello_20_3.0.png - text: waving hello output: url: images/waving hello_21_3.0.png - text: waving hello output: url: images/waving hello_22_3.0.png tags: - text-to-image - stable-diffusion-xl - lora - template:sd-lora - template:sdxl-lora - sdxl-sliders - ntcai.xyz-sliders - concept - diffusers license: "mit" inference: false instance_prompt: "waving hello" base_model: "stabilityai/stable-diffusion-xl-base-1.0" --- # ntcai.xyz slider - waving hello (SDXL LoRA) \| Strength: -3 \| Strength: 0 \| Strength: 3 \| \| --- \| --- \| --- \| \| <img src="images/waving hello_17_-3.0.png" width=256 height=256 /> \| <img src="images/waving hello_17_0.0.png" width=256 height=256 /> \| <img src="images/waving hello_17_3.0.png" width=256 height=256 /> \| \| <img src="images/waving hello_19_-3.0.png" width=256 height=256 /> \| <img src="images/waving hello_19_0.0.png" width=256 height=256 /> \| <img src="images/waving hello_19_3.0.png" width=256 height=256 /> \| \| <img src="images/waving hello_20_-3.0.png" width=256 height=256 /> \| <img src="images/waving hello_20_0.0.png" width=256 height=256 /> \| <img src="images/waving hello_20_3.0.png" width=256 height=256 /> \| ## Download Weights for this model are available in Safetensors format. ## Trigger words You can apply this LoRA with trigger words for additional effect: ``` waving hello ``` ## Use in diffusers ```python from diffusers import StableDiffusionXLPipeline from diffusers import EulerAncestralDiscreteScheduler import torch pipe = StableDiffusionXLPipeline.from_single_file("https://huggingface.co/martyn/sdxl-turbo-mario-merge-top-rated/blob/main/topRatedTurboxlLCM_v10.safetensors") pipe.to("cuda") pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config) # Load the LoRA pipe.load_lora_weights('ntc-ai/SDXL-LoRA-slider.waving-hello', weight_name='waving hello.safetensors', adapter_name="waving hello") # Activate the LoRA pipe.set_adapters(["waving hello"], adapter_weights=[2.0]) prompt = "medieval rich kingpin sitting in a tavern, waving hello" negative_prompt = "nsfw" width = 512 height = 512 num_inference_steps = 10 guidance_scale = 2 image = pipe(prompt, negative_prompt=negative_prompt, width=width, height=height, guidance_scale=guidance_scale, num_inference_steps=num_inference_steps).images[0] image.save('result.png') ``` ## Support the Patreon If you like this model please consider [joining our Patreon](https://www.patreon.com/NTCAI). By joining our Patreon, you'll gain access to an ever-growing library of over 1070+ unique and diverse LoRAs, covering a wide range of styles and genres. You'll also receive early access to new models and updates, exclusive behind-the-scenes content, and the powerful LoRA slider creator, allowing you to craft your own custom LoRAs and experiment with endless possibilities. Your support on Patreon will allow us to continue developing and refining new models. ## Other resources - [CivitAI](https://civitai.com/user/ntc) - Follow ntc on Civit for even more LoRAs - [ntcai.xyz](https://ntcai.xyz) - See ntcai.xyz to find more articles and LoRAs
yihang7/zephyr-7b-dpo-full-hydrox-safe	yihang7	2024-01-13T03:21:02Z	6	0	transformers	[ "transformers", "tensorboard", "safetensors", "mistral", "text-generation", "generated_from_trainer", "conversational", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T00:50:14Z	--- tags: - generated_from_trainer model-index: - name: zephyr-7b-dpo-full-hydrox-safe results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # zephyr-7b-dpo-full-hydrox-safe This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.0018 - Rewards/chosen: 1.5348 - Rewards/rejected: -41.0322 - Rewards/accuracies: 0.9992 - Rewards/margins: 42.5670 - Logps/rejected: -855.1025 - Logps/chosen: -270.7936 - Logits/rejected: -2.8721 - Logits/chosen: -3.0906 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-07 - train_batch_size: 8 - eval_batch_size: 4 - seed: 42 - distributed_type: multi-GPU - num_devices: 8 - total_train_batch_size: 64 - total_eval_batch_size: 32 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 3 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Rewards/chosen \| Rewards/rejected \| Rewards/accuracies \| Rewards/margins \| Logps/rejected \| Logps/chosen \| Logits/rejected \| Logits/chosen \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:--------------:\|:----------------:\|:------------------:\|:---------------:\|:--------------:\|:------------:\|:---------------:\|:-------------:\| \| 0.201 \| 0.03 \| 100 \| 0.1790 \| 0.8766 \| -1.8347 \| 0.9571 \| 2.7113 \| -463.1273 \| -277.3755 \| -2.9298 \| -3.0062 \| \| 0.1124 \| 0.07 \| 200 \| 0.0931 \| 1.4748 \| -3.5536 \| 0.9764 \| 5.0284 \| -480.3171 \| -271.3938 \| -2.9144 \| -2.9984 \| \| 0.0657 \| 0.1 \| 300 \| 0.1207 \| 1.7581 \| -4.7603 \| 0.9806 \| 6.5185 \| -492.3842 \| -268.5603 \| -2.9028 \| -2.9908 \| \| 0.0676 \| 0.14 \| 400 \| 0.0446 \| 1.3392 \| -8.0193 \| 0.9857 \| 9.3585 \| -524.9738 \| -272.75 \| -2.9171 \| -3.0132 \| \| 0.0477 \| 0.17 \| 500 \| 0.0581 \| 1.3777 \| -9.9106 \| 0.9882 \| 11.2883 \| -543.8868 \| -272.3650 \| -2.8810 \| -2.9695 \| \| 0.0404 \| 0.2 \| 600 \| 0.0845 \| 0.3594 \| -12.1256 \| 0.9840 \| 12.4850 \| -566.0372 \| -282.5481 \| -2.8164 \| -2.9235 \| \| 0.0122 \| 0.24 \| 700 \| 0.0831 \| 0.4374 \| -19.3178 \| 0.9848 \| 19.7552 \| -637.9587 \| -281.7681 \| -2.7024 \| -2.8110 \| \| 0.0183 \| 0.27 \| 800 \| 0.0775 \| -0.0400 \| -18.4153 \| 0.9891 \| 18.3753 \| -628.9333 \| -286.5414 \| -2.8078 \| -2.9358 \| \| 0.0555 \| 0.31 \| 900 \| 0.0589 \| 0.4664 \| -23.4223 \| 0.9924 \| 23.8887 \| -679.0039 \| -281.4781 \| -2.8207 \| -2.9406 \| \| 0.0376 \| 0.34 \| 1000 \| 0.0552 \| 0.6445 \| -26.2370 \| 0.9874 \| 26.8814 \| -707.1502 \| -279.6971 \| -2.7989 \| -2.9241 \| \| 0.0438 \| 0.37 \| 1100 \| 0.0685 \| 1.7090 \| -23.4352 \| 0.9874 \| 25.1442 \| -679.1324 \| -269.0518 \| -2.8438 \| -3.0501 \| \| 0.074 \| 0.41 \| 1200 \| 0.0715 \| 1.4961 \| -24.8055 \| 0.9882 \| 26.3016 \| -692.8359 \| -271.1810 \| -2.7269 \| -2.8728 \| \| 0.1029 \| 0.44 \| 1300 \| 0.1098 \| 0.9065 \| -28.0158 \| 0.9882 \| 28.9223 \| -724.9388 \| -277.0765 \| -2.7411 \| -2.9059 \| \| 0.0271 \| 0.48 \| 1400 \| 0.0346 \| 0.7923 \| -26.9632 \| 0.9907 \| 27.7555 \| -714.4124 \| -278.2186 \| -2.7555 \| -2.9286 \| \| 0.0219 \| 0.51 \| 1500 \| 0.0405 \| 0.5053 \| -27.0039 \| 0.9933 \| 27.5092 \| -714.8198 \| -281.0890 \| -2.7955 \| -2.9553 \| \| 0.0214 \| 0.54 \| 1600 \| 0.0293 \| 0.7598 \| -22.6033 \| 0.9941 \| 23.3631 \| -670.8132 \| -278.5433 \| -2.7650 \| -2.9211 \| \| 0.0221 \| 0.58 \| 1700 \| 0.0206 \| 0.6463 \| -25.7837 \| 0.9949 \| 26.4300 \| -702.6176 \| -279.6782 \| -2.7340 \| -2.9148 \| \| 0.0339 \| 0.61 \| 1800 \| 0.0204 \| 1.2579 \| -22.2971 \| 0.9966 \| 23.5550 \| -667.7516 \| -273.5627 \| -2.7337 \| -2.8868 \| \| 0.0662 \| 0.65 \| 1900 \| 0.0290 \| 0.9343 \| -30.7737 \| 0.9975 \| 31.7081 \| -752.5181 \| -276.7985 \| -2.7130 \| -2.8608 \| \| 0.1032 \| 0.68 \| 2000 \| 0.0358 \| 1.6616 \| -22.3776 \| 0.9975 \| 24.0392 \| -668.5566 \| -269.5255 \| -2.7454 \| -2.8772 \| \| 0.0059 \| 0.71 \| 2100 \| 0.0148 \| 0.9711 \| -28.7176 \| 0.9949 \| 29.6887 \| -731.9562 \| -276.4307 \| -2.6758 \| -2.8253 \| \| 0.0096 \| 0.75 \| 2200 \| 0.0282 \| 0.9406 \| -25.7753 \| 0.9949 \| 26.7158 \| -702.5333 \| -276.7361 \| -2.9417 \| -3.1166 \| \| 0.0117 \| 0.78 \| 2300 \| 0.0489 \| -0.3402 \| -33.3639 \| 0.9924 \| 33.0238 \| -778.4199 \| -289.5432 \| -3.0059 \| -3.1712 \| \| 0.0034 \| 0.82 \| 2400 \| 0.0136 \| 1.4644 \| -27.4436 \| 0.9966 \| 28.9079 \| -719.2164 \| -271.4980 \| -2.6572 \| -2.7758 \| \| 0.0448 \| 0.85 \| 2500 \| 0.0122 \| -0.7904 \| -29.3707 \| 0.9983 \| 28.5803 \| -738.4877 \| -294.0459 \| -2.7608 \| -2.8870 \| \| 0.0606 \| 0.88 \| 2600 \| 0.0723 \| 0.6824 \| -40.6246 \| 0.9949 \| 41.3070 \| -851.0272 \| -279.3180 \| -2.6804 \| -2.8281 \| \| 0.0172 \| 0.92 \| 2700 \| 0.0151 \| 1.7774 \| -28.1840 \| 0.9958 \| 29.9614 \| -726.6205 \| -268.3677 \| -2.7720 \| -2.9085 \| \| 0.0206 \| 0.95 \| 2800 \| 0.0105 \| 0.0966 \| -39.5676 \| 0.9975 \| 39.6643 \| -840.4568 \| -285.1752 \| -2.8397 \| -2.9803 \| \| 0.0093 \| 0.99 \| 2900 \| 0.0165 \| 0.6995 \| -33.4341 \| 0.9966 \| 34.1336 \| -779.1221 \| -279.1465 \| -2.8086 \| -2.9511 \| \| 0.0008 \| 1.02 \| 3000 \| 0.0101 \| -0.0075 \| -36.2959 \| 0.9983 \| 36.2884 \| -807.7395 \| -286.2166 \| -2.9213 \| -3.0591 \| \| 0.0015 \| 1.05 \| 3100 \| 0.0091 \| -0.2430 \| -38.1496 \| 0.9966 \| 37.9066 \| -826.2767 \| -288.5718 \| -2.8160 \| -2.9881 \| \| 0.001 \| 1.09 \| 3200 \| 0.0069 \| 0.7306 \| -37.6657 \| 0.9966 \| 38.3964 \| -821.4379 \| -278.8353 \| -2.8287 \| -2.9686 \| \| 0.0447 \| 1.12 \| 3300 \| 0.0148 \| 0.4641 \| -39.1245 \| 0.9975 \| 39.5886 \| -836.0254 \| -281.5004 \| -2.8502 \| -2.9821 \| \| 0.0015 \| 1.16 \| 3400 \| 0.0145 \| 1.2867 \| -34.9331 \| 0.9983 \| 36.2198 \| -794.1116 \| -273.2742 \| -2.8645 \| -2.9793 \| \| 0.0069 \| 1.19 \| 3500 \| 0.0120 \| 1.0980 \| -33.3648 \| 0.9983 \| 34.4628 \| -778.4290 \| -275.1619 \| -2.8298 \| -2.9435 \| \| 0.0006 \| 1.22 \| 3600 \| 0.0144 \| 0.4168 \| -38.9632 \| 0.9975 \| 39.3800 \| -834.4130 \| -281.9736 \| -2.8061 \| -2.9050 \| \| 0.0532 \| 1.26 \| 3700 \| 0.0163 \| 1.2493 \| -35.1549 \| 0.9983 \| 36.4042 \| -796.3295 \| -273.6483 \| -3.0117 \| -3.1271 \| \| 0.0041 \| 1.29 \| 3800 \| 0.0138 \| 0.6365 \| -36.2498 \| 0.9966 \| 36.8863 \| -807.2789 \| -279.7767 \| -2.9031 \| -3.0283 \| \| 0.0043 \| 1.33 \| 3900 \| 0.0095 \| 0.8546 \| -36.5756 \| 0.9992 \| 37.4302 \| -810.5366 \| -277.5952 \| -2.9713 \| -3.0945 \| \| 0.0016 \| 1.36 \| 4000 \| 0.0098 \| 0.4731 \| -41.9454 \| 0.9992 \| 42.4186 \| -864.2352 \| -281.4105 \| -3.0390 \| -3.1458 \| \| 0.0018 \| 1.39 \| 4100 \| 0.0103 \| 1.1650 \| -31.5836 \| 0.9992 \| 32.7487 \| -760.6172 \| -274.4915 \| -3.0057 \| -3.1525 \| \| 0.0015 \| 1.43 \| 4200 \| 0.0097 \| 0.7472 \| -35.6321 \| 0.9992 \| 36.3793 \| -801.1016 \| -278.6696 \| -2.9895 \| -3.1466 \| \| 0.0078 \| 1.46 \| 4300 \| 0.0171 \| 0.4428 \| -47.4943 \| 0.9983 \| 47.9370 \| -919.7234 \| -281.7140 \| -3.0189 \| -3.1577 \| \| 0.0006 \| 1.5 \| 4400 \| 0.0117 \| 2.3601 \| -36.7779 \| 0.9992 \| 39.1379 \| -812.5596 \| -262.5412 \| -2.9504 \| -3.0998 \| \| 0.0021 \| 1.53 \| 4500 \| 0.0096 \| 1.8317 \| -33.0472 \| 0.9983 \| 34.8789 \| -775.2529 \| -267.8244 \| -2.8650 \| -3.0361 \| \| 0.0022 \| 1.56 \| 4600 \| 0.0124 \| 1.8244 \| -35.2193 \| 0.9983 \| 37.0437 \| -796.9737 \| -267.8973 \| -2.8983 \| -3.0580 \| \| 0.0025 \| 1.6 \| 4700 \| 0.0120 \| 1.4261 \| -35.4094 \| 0.9983 \| 36.8355 \| -798.8742 \| -271.8802 \| -2.8938 \| -3.0468 \| \| 0.0154 \| 1.63 \| 4800 \| 0.0070 \| 0.7832 \| -39.4533 \| 0.9983 \| 40.2365 \| -839.3135 \| -278.3098 \| -2.9491 \| -3.1257 \| \| 0.0025 \| 1.67 \| 4900 \| 0.0070 \| 0.5157 \| -40.3519 \| 0.9983 \| 40.8676 \| -848.2994 \| -280.9843 \| -2.8537 \| -3.0217 \| \| 0.0022 \| 1.7 \| 5000 \| 0.0060 \| 1.8378 \| -36.1887 \| 0.9983 \| 38.0265 \| -806.6680 \| -267.7638 \| -2.8008 \| -2.9944 \| \| 0.0116 \| 1.73 \| 5100 \| 0.0067 \| 1.6571 \| -30.2966 \| 0.9983 \| 31.9537 \| -747.7468 \| -269.5709 \| -2.8245 \| -3.0273 \| \| 0.0027 \| 1.77 \| 5200 \| 0.0056 \| 1.8559 \| -28.9170 \| 0.9983 \| 30.7729 \| -733.9506 \| -267.5823 \| -2.8308 \| -3.0333 \| \| 0.0008 \| 1.8 \| 5300 \| 0.0120 \| 0.1806 \| -39.8991 \| 0.9983 \| 40.0797 \| -843.7715 \| -284.3351 \| -3.0642 \| -3.2708 \| \| 0.0008 \| 1.84 \| 5400 \| 0.0042 \| 1.8857 \| -32.3213 \| 0.9983 \| 34.2070 \| -767.9935 \| -267.2849 \| -2.9745 \| -3.1835 \| \| 0.0419 \| 1.87 \| 5500 \| 0.0049 \| 0.4195 \| -42.5787 \| 0.9983 \| 42.9982 \| -870.5677 \| -281.9464 \| -2.9390 \| -3.1435 \| \| 0.0055 \| 1.9 \| 5600 \| 0.0039 \| 0.8564 \| -39.5863 \| 0.9966 \| 40.4428 \| -840.6442 \| -277.5772 \| -2.9517 \| -3.1465 \| \| 0.0016 \| 1.94 \| 5700 \| 0.0039 \| 1.5280 \| -38.4427 \| 0.9983 \| 39.9707 \| -829.2079 \| -270.8615 \| -2.9334 \| -3.1210 \| \| 0.0015 \| 1.97 \| 5800 \| 0.0031 \| 2.5689 \| -32.1821 \| 0.9992 \| 34.7510 \| -766.6016 \| -260.4530 \| -2.9266 \| -3.1170 \| \| 0.0022 \| 2.01 \| 5900 \| 0.0154 \| 1.5272 \| -38.5973 \| 0.9983 \| 40.1246 \| -830.7538 \| -270.8692 \| -2.9905 \| -3.1795 \| \| 0.0001 \| 2.04 \| 6000 \| 0.0153 \| 1.4304 \| -38.7250 \| 0.9983 \| 40.1554 \| -832.0308 \| -271.8378 \| -3.0212 \| -3.2078 \| \| 0.0004 \| 2.07 \| 6100 \| 0.0151 \| 0.9974 \| -40.5361 \| 0.9983 \| 41.5335 \| -850.1418 \| -276.1674 \| -3.0092 \| -3.1920 \| \| 0.0032 \| 2.11 \| 6200 \| 0.0143 \| 2.2246 \| -35.4872 \| 0.9983 \| 37.7117 \| -799.6525 \| -263.8961 \| -3.0169 \| -3.2017 \| \| 0.0026 \| 2.14 \| 6300 \| 0.0138 \| 2.1149 \| -36.0655 \| 0.9983 \| 38.1804 \| -805.4358 \| -264.9924 \| -3.0121 \| -3.1974 \| \| 0.0052 \| 2.18 \| 6400 \| 0.0093 \| 2.0832 \| -38.0872 \| 0.9983 \| 40.1704 \| -825.6531 \| -265.3097 \| -2.9119 \| -3.1190 \| \| 0.0002 \| 2.21 \| 6500 \| 0.0031 \| 2.4808 \| -33.4480 \| 0.9992 \| 35.9287 \| -779.2603 \| -261.3340 \| -2.9056 \| -3.1235 \| \| 0.0014 \| 2.24 \| 6600 \| 0.0029 \| 2.1622 \| -35.1487 \| 0.9992 \| 37.3110 \| -796.2680 \| -264.5192 \| -2.9322 \| -3.1505 \| \| 0.0001 \| 2.28 \| 6700 \| 0.0029 \| 2.6987 \| -34.3087 \| 0.9992 \| 37.0074 \| -787.8676 \| -259.1545 \| -2.9528 \| -3.1683 \| \| 0.0004 \| 2.31 \| 6800 \| 0.0028 \| 2.2809 \| -36.5033 \| 0.9992 \| 38.7842 \| -809.8135 \| -263.3324 \| -2.9377 \| -3.1456 \| \| 0.0268 \| 2.35 \| 6900 \| 0.0028 \| 1.9206 \| -37.6915 \| 0.9983 \| 39.6121 \| -821.6957 \| -266.9360 \| -2.9110 \| -3.1199 \| \| 0.0012 \| 2.38 \| 7000 \| 0.0026 \| 2.0357 \| -34.8535 \| 0.9992 \| 36.8892 \| -793.3152 \| -265.7842 \| -2.9165 \| -3.1204 \| \| 0.0002 \| 2.41 \| 7100 \| 0.0031 \| 0.4887 \| -42.0925 \| 0.9983 \| 42.5812 \| -865.7061 \| -281.2549 \| -2.8696 \| -3.0815 \| \| 0.0006 \| 2.45 \| 7200 \| 0.0027 \| 1.6025 \| -38.2160 \| 0.9992 \| 39.8185 \| -826.9409 \| -270.1166 \| -2.8921 \| -3.1009 \| \| 0.0037 \| 2.48 \| 7300 \| 0.0027 \| 1.2715 \| -42.6264 \| 0.9992 \| 43.8978 \| -871.0444 \| -273.4270 \| -2.8870 \| -3.0973 \| \| 0.0023 \| 2.52 \| 7400 \| 0.0026 \| 1.3232 \| -42.5318 \| 0.9992 \| 43.8549 \| -870.0981 \| -272.9100 \| -2.8912 \| -3.1000 \| \| 0.0012 \| 2.55 \| 7500 \| 0.0026 \| 1.4663 \| -41.9917 \| 0.9992 \| 43.4581 \| -864.6981 \| -271.4783 \| -2.8990 \| -3.1084 \| \| 0.0012 \| 2.58 \| 7600 \| 0.0025 \| 1.2190 \| -42.9826 \| 0.9992 \| 44.2016 \| -874.6072 \| -273.9521 \| -2.8991 \| -3.1070 \| \| 0.0001 \| 2.62 \| 7700 \| 0.0026 \| 1.7469 \| -40.1663 \| 0.9992 \| 41.9132 \| -846.4437 \| -268.6728 \| -2.8615 \| -3.0744 \| \| 0.0001 \| 2.65 \| 7800 \| 0.0026 \| 1.7291 \| -40.4524 \| 0.9992 \| 42.1815 \| -849.3045 \| -268.8504 \| -2.8619 \| -3.0751 \| \| 0.003 \| 2.69 \| 7900 \| 0.0018 \| 1.8715 \| -39.4482 \| 0.9992 \| 41.3197 \| -839.2630 \| -267.4267 \| -2.8391 \| -3.0504 \| \| 0.0001 \| 2.72 \| 8000 \| 0.0018 \| 1.4707 \| -42.0026 \| 0.9992 \| 43.4733 \| -864.8069 \| -271.4347 \| -2.8587 \| -3.0725 \| \| 0.0001 \| 2.75 \| 8100 \| 0.0018 \| 1.7032 \| -41.2255 \| 0.9992 \| 42.9287 \| -857.0361 \| -269.1100 \| -2.8611 \| -3.0771 \| \| 0.0001 \| 2.79 \| 8200 \| 0.0018 \| 1.7053 \| -41.5715 \| 0.9992 \| 43.2768 \| -860.4957 \| -269.0886 \| -2.8608 \| -3.0769 \| \| 0.0 \| 2.82 \| 8300 \| 0.0017 \| 1.5198 \| -41.7233 \| 0.9992 \| 43.2431 \| -862.0136 \| -270.9432 \| -2.8563 \| -3.0742 \| \| 0.0 \| 2.86 \| 8400 \| 0.0018 \| 1.7544 \| -40.1153 \| 0.9992 \| 41.8696 \| -845.9333 \| -268.5981 \| -2.8694 \| -3.0853 \| \| 0.0003 \| 2.89 \| 8500 \| 0.0018 \| 1.6997 \| -40.2116 \| 0.9992 \| 41.9113 \| -846.8971 \| -269.1451 \| -2.8782 \| -3.0951 \| \| 0.0071 \| 2.92 \| 8600 \| 0.0018 \| 1.5913 \| -40.7278 \| 0.9992 \| 42.3191 \| -852.0585 \| -270.2284 \| -2.8728 \| -3.0908 \| \| 0.0013 \| 2.96 \| 8700 \| 0.0018 \| 1.5263 \| -41.0232 \| 0.9992 \| 42.5495 \| -855.0129 \| -270.8784 \| -2.8716 \| -3.0900 \| \| 0.0003 \| 2.99 \| 8800 \| 0.0019 \| 1.5305 \| -41.0444 \| 0.9992 \| 42.5748 \| -855.2242 \| -270.8367 \| -2.8724 \| -3.0907 \| ### Framework versions - Transformers 4.35.0 - Pytorch 2.1.1+cu121 - Datasets 2.14.6 - Tokenizers 0.14.1
shidowake/mistral-base-bnb-4bit-alpaca-lora-adaptor	shidowake	2024-01-13T03:17:30Z	0	0	peft	[ "peft", "safetensors", "arxiv:1910.09700", "base_model:unsloth/mistral-7b", "base_model:adapter:unsloth/mistral-7b", "region:us" ]	null	2024-01-13T03:17:21Z	--- library_name: peft base_model: unsloth/mistral-7b --- # Model Card for Model ID <!-- Provide a quick summary of what the model is/does. --> ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> - Developed by: [More Information Needed] - Funded by [optional]: [More Information Needed] - Shared by [optional]: [More Information Needed] - Model type: [More Information Needed] - Language(s) (NLP): [More Information Needed] - License: [More Information Needed] - Finetuned from model [optional]: [More Information Needed] ### Model Sources [optional] <!-- Provide the basic links for the model. --> - Repository: [More Information Needed] - Paper [optional]: [More Information Needed] - Demo [optional]: [More Information Needed] ## Uses <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> [More Information Needed] ### Downstream Use [optional] <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app --> [More Information Needed] ### Out-of-Scope Use <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> [More Information Needed] ## Bias, Risks, and Limitations <!-- This section is meant to convey both technical and sociotechnical limitations. --> [More Information Needed] ### Recommendations <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. --> Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. ## How to Get Started with the Model Use the code below to get started with the model. [More Information Needed] ## Training Details ### Training Data <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> [More Information Needed] ### Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> #### Preprocessing [optional] [More Information Needed] #### Training Hyperparameters - Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision --> #### Speeds, Sizes, Times [optional] <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. --> [More Information Needed] ## Evaluation <!-- This section describes the evaluation protocols and provides the results. --> ### Testing Data, Factors & Metrics #### Testing Data <!-- This should link to a Dataset Card if possible. --> [More Information Needed] #### Factors <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. --> [More Information Needed] #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> [More Information Needed] ### Results [More Information Needed] #### Summary ## Model Examination [optional] <!-- Relevant interpretability work for the model goes here --> [More Information Needed] ## Environmental Impact <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly --> Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - Hardware Type: [More Information Needed] - Hours used: [More Information Needed] - Cloud Provider: [More Information Needed] - Compute Region: [More Information Needed] - Carbon Emitted: [More Information Needed] ## Technical Specifications [optional] ### Model Architecture and Objective [More Information Needed] ### Compute Infrastructure [More Information Needed] #### Hardware [More Information Needed] #### Software [More Information Needed] ## Citation [optional] <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> BibTeX: [More Information Needed] APA: [More Information Needed] ## Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> [More Information Needed] ## More Information [optional] [More Information Needed] ## Model Card Authors [optional] [More Information Needed] ## Model Card Contact [More Information Needed] ### Framework versions - PEFT 0.7.1
venkycs/Zyte-1B-gguf	venkycs	2024-01-13T03:06:42Z	5	1	null	[ "gguf", "tinyllama", "leader", "dpo", "chat", "en", "license:apache-2.0", "endpoints_compatible", "region:us", "conversational" ]	null	2024-01-12T13:39:26Z	--- license: apache-2.0 language: - en metrics: - accuracy tags: - gguf - tinyllama - leader - dpo - chat --- # Model Card for Model ID Here is original model https://huggingface.co/aihub-app/zyte-1B LM Studio - https://lmstudio.ai (Fully Supported, make sure you select Zypher template format for inference with model.) We have chosen to upload our model exclusively in the Q8 format, taking into account its compact size. This decision strikes an optimal balance between speed, performance, and memory usage, ensuring efficient and effective utilization of resources. ## Model Card Authors Special thanks to Venkatesh Siddi for his invaluable support in the conversion of the model to the GGUF format. His expertise and assistance have been instrumental in this process. Connect with him on LinkedIn: https://www.linkedin.com/in/venkycs/
yangzhou301/poca-SoccerTwos	yangzhou301	2024-01-13T02:12:39Z	12	0	ml-agents	[ "ml-agents", "tensorboard", "onnx", "SoccerTwos", "deep-reinforcement-learning", "reinforcement-learning", "ML-Agents-SoccerTwos", "region:us" ]	reinforcement-learning	2024-01-13T02:12:28Z	--- library_name: ml-agents tags: - SoccerTwos - deep-reinforcement-learning - reinforcement-learning - ML-Agents-SoccerTwos --- # poca Agent playing SoccerTwos This is a trained model of a poca agent playing SoccerTwos using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents). ## Usage (with ML-Agents) The Documentation: https://unity-technologies.github.io/ml-agents/ML-Agents-Toolkit-Documentation/ We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub: - A short tutorial where you teach Huggy the Dog 🐶 to fetch the stick and then play with him directly in your browser: https://huggingface.co/learn/deep-rl-course/unitbonus1/introduction - A longer tutorial to understand how works ML-Agents: https://huggingface.co/learn/deep-rl-course/unit5/introduction ### Resume the training ```bash mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume ``` ### Watch your Agent play You can watch your agent playing directly in your browser 1. If the environment is part of ML-Agents official environments, go to https://huggingface.co/unity 2. Step 1: Find your model_id: yangzhou301/poca-SoccerTwos 3. Step 2: Select your .nn /.onnx file 4. Click on Watch the agent play 👀
Luckythespacecat/Somethin	Luckythespacecat	2024-01-13T01:38:16Z	0	0	null	[ "text-to-image", "region:us" ]	text-to-image	2024-01-13T01:36:41Z	--- pipeline_tag: text-to-image ---
littlepanic1234/tmp_trainer	littlepanic1234	2024-01-13T01:35:10Z	0	0	null	[ "safetensors", "generated_from_trainer", "base_model:mistralai/Mistral-7B-v0.1", "base_model:finetune:mistralai/Mistral-7B-v0.1", "license:apache-2.0", "region:us" ]	null	2024-01-13T01:33:33Z	--- license: apache-2.0 base_model: mistralai/Mistral-7B-v0.1 tags: - generated_from_trainer model-index: - name: tmp_trainer results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # tmp_trainer This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3.0 ### Framework versions - Transformers 4.36.2 - Pytorch 2.1.2+cu121 - Datasets 2.14.1 - Tokenizers 0.15.0
prashrex/fintuned_model	prashrex	2024-01-13T01:33:26Z	4	0	peft	[ "peft", "safetensors", "arxiv:1910.09700", "base_model:TinyLlama/TinyLlama-1.1B-Chat-v1.0", "base_model:adapter:TinyLlama/TinyLlama-1.1B-Chat-v1.0", "region:us" ]	null	2024-01-12T17:39:16Z	--- library_name: peft base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0 --- # Model Card for Model ID <!-- Provide a quick summary of what the model is/does. --> ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> - Developed by: [More Information Needed] - Funded by [optional]: [More Information Needed] - Shared by [optional]: [More Information Needed] - Model type: [More Information Needed] - Language(s) (NLP): [More Information Needed] - License: [More Information Needed] - Finetuned from model [optional]: [More Information Needed] ### Model Sources [optional] <!-- Provide the basic links for the model. --> - Repository: [More Information Needed] - Paper [optional]: [More Information Needed] - Demo [optional]: [More Information Needed] ## Uses <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> [More Information Needed] ### Downstream Use [optional] <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app --> [More Information Needed] ### Out-of-Scope Use <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> [More Information Needed] ## Bias, Risks, and Limitations <!-- This section is meant to convey both technical and sociotechnical limitations. --> [More Information Needed] ### Recommendations <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. --> Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. ## How to Get Started with the Model Use the code below to get started with the model. [More Information Needed] ## Training Details ### Training Data <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> [More Information Needed] ### Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> #### Preprocessing [optional] [More Information Needed] #### Training Hyperparameters - Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision --> #### Speeds, Sizes, Times [optional] <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. --> [More Information Needed] ## Evaluation <!-- This section describes the evaluation protocols and provides the results. --> ### Testing Data, Factors & Metrics #### Testing Data <!-- This should link to a Dataset Card if possible. --> [More Information Needed] #### Factors <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. --> [More Information Needed] #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> [More Information Needed] ### Results [More Information Needed] #### Summary ## Model Examination [optional] <!-- Relevant interpretability work for the model goes here --> [More Information Needed] ## Environmental Impact <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly --> Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - Hardware Type: [More Information Needed] - Hours used: [More Information Needed] - Cloud Provider: [More Information Needed] - Compute Region: [More Information Needed] - Carbon Emitted: [More Information Needed] ## Technical Specifications [optional] ### Model Architecture and Objective [More Information Needed] ### Compute Infrastructure [More Information Needed] #### Hardware [More Information Needed] #### Software [More Information Needed] ## Citation [optional] <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> BibTeX: [More Information Needed] APA: [More Information Needed] ## Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> [More Information Needed] ## More Information [optional] [More Information Needed] ## Model Card Authors [optional] [More Information Needed] ## Model Card Contact [More Information Needed] ### Framework versions - PEFT 0.7.2.dev0
cnatale/mistral_7b_instruct_v0_1_txt_2_sql	cnatale	2024-01-13T01:32:30Z	1	0	peft	[ "peft", "tensorboard", "safetensors", "trl", "sft", "generated_from_trainer", "dataset:generator", "base_model:mistralai/Mistral-7B-Instruct-v0.1", "base_model:adapter:mistralai/Mistral-7B-Instruct-v0.1", "license:apache-2.0", "region:us" ]	null	2024-01-13T00:49:27Z	--- license: apache-2.0 library_name: peft tags: - trl - sft - generated_from_trainer datasets: - generator base_model: mistralai/Mistral-7B-Instruct-v0.1 model-index: - name: mistral_7b_instruct_v0_1_txt_2_sql results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # mistral_7b_instruct_v0_1_txt_2_sql This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the generator dataset. It achieves the following results on the evaluation set: - Loss: 0.6482 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001 - train_batch_size: 4 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: constant - lr_scheduler_warmup_steps: 0.03 - training_steps: 80 - mixed_precision_training: Native AMP ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| 1.3599 \| 0.71 \| 10 \| 1.0764 \| \| 1.0005 \| 1.43 \| 20 \| 0.8726 \| \| 0.8508 \| 2.14 \| 30 \| 0.7721 \| \| 0.7585 \| 2.86 \| 40 \| 0.7218 \| \| 0.7078 \| 3.57 \| 50 \| 0.6825 \| \| 0.6413 \| 4.29 \| 60 \| 0.6594 \| \| 0.6128 \| 5.0 \| 70 \| 0.6508 \| \| 0.5745 \| 5.71 \| 80 \| 0.6482 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.1.2+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
audreyt/Breeze-7B-Base-v0.1-GGUF	audreyt	2024-01-13T01:19:32Z	66	6	transformers	[ "transformers", "gguf", "text-generation", "zh", "license:apache-2.0", "region:us" ]	text-generation	2024-01-13T00:22:30Z	--- license: apache-2.0 language: - zh library_name: transformers pipeline_tag: text-generation inference: false quantized_by: audreyt --- # Breeze-7B-Base-v0.1-GGUF - Model creator: [MediaTek Research](https://huggingface.co/MediaTek-Research) - Original model: [Breeze-7B-Base-v0.1](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v0.1) ## Description This repo contains GGUF format model files for MediaTek Research's [Breeze-7B-Base-v0.1](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v0.1). <!-- README_GGUF.md-about-gguf start --> ### About GGUF GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp. Here is an incomplete list of clients and libraries that are known to support GGUF: * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option. * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration. * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling. * [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel. * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023. * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection. * [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration. * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server. * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use. * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models. <!-- README_GGUF.md-about-gguf end --> # Original model card Breeze-7B is a language model family that builds on top of [Mistral-7B](https://huggingface.co/mistralai/Mistral-7B-v0.1), specifically intended for Traditional Chinese use. [Breeze-7B-Base](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v0.1) is the base model for the Breeze-7B series. It is suitable for use if you have substantial fine-tuning data to tune it for your specific use case. [Breeze-7B-Instruct](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-v0.1) derives from the base model Breeze-7B-Base, making the resulting model amenable to be used as-is for commonly seen tasks. [Breeze-7B-Instruct-64k](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-64k-v0.1) is a slightly modified version of Breeze-7B-Instruct to enable a 64k-token context length. Roughly speaking, that is equivalent to 88k Traditional Chinese characters. The current release version of Breeze-7B is v0.1. Practicality-wise: - Breeze-7B-Base expands the original vocabulary with additional 30,000 Traditional Chinese tokens. With the expanded vocabulary, everything else being equal, Breeze-7B operates at twice the inference speed for Traditional Chinese to Mistral-7B and Llama 7B. [See [Inference Performance](#inference-performance).] - Breeze-7B-Instruct can be used as is for common tasks such as Q&A, RAG, multi-round chat, and summarization. - In particular, Breeze-7B-Instruct-64k can perform tasks at a document level, not a chapter level. Performance-wise: - Breeze-7B-Instruct demonstrates impressive performance in benchmarks for Traditional Chinese, when compared to similar sized open-source contemporaries such as Taiwan-LLM-7B/13B-chat, QWen-7B-Chat, and Yi-6B-Chat. [See [Chat Model Performance](#chat-model-performance).] - Breeze-7B-Instruct shows comparable results to Mistral-7B-Instruct-v0.1 on the MMLU and MT-Bench benchmarks. [See [Chat Model Performance](#chat-model-performance).] A project by the members (in alphabetical order): Chan-Jan Hsu 許湛然, Chang-Le Liu 劉昶樂, Feng-Ting Liao 廖峰挺, Po-Chun Hsu 許博竣, Yi-Chang Chen 陳宜昌, and the supervisor Da-Shan Shiu 許大山. ## Features - Breeze-7B-Base-v0.1 - Expanding the vocabulary dictionary size from 32k to 62k to better support Traditional Chinese - 8k-token context length - Breeze-7B-Instruct-v0.1 - Expanding the vocabulary dictionary size from 32k to 62k to better support Traditional Chinese - 8k-token context length - Multi-turn dialogue (without special handling for harmfulness) - Breeze-7B-Instruct-64k-v0.1 - Expanding the vocabulary dictionary size from 32k to 62k to better support Traditional Chinese - 64k-token context length - Multi-turn dialogue (without special handling for harmfulness) ## Model Details - Breeze-7B-Base-v0.1 - Finetuned from: [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) - Model type: Causal decoder-only transformer language model - Language: English and Traditional Chinese (zh-tw) - Breeze-7B-Instruct-v0.1 - Finetuned from: [MediaTek-Research/Breeze-7B-Base-v0.1](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v0.1) - Model type: Causal decoder-only transformer language model - Language: English and Traditional Chinese (zh-tw) - Breeze-7B-Instruct-64k-v0.1 - Finetuned from: [MediaTek-Research/Breeze-7B-Instruct-v0.1](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-v0.1) - Model type: Causal decoder-only transformer language model - Language: English and Traditional Chinese (zh-tw) ## Base Model Performance TMMLU+, DRCD, and Table source from [MediaTek-Research/TCEval-v2](https://huggingface.co/datasets/MediaTek-Research/TCEval-v2). [MediaTek-Research/TCEval-v2](https://huggingface.co/datasets/MediaTek-Research/TCEval-v2) derives from [TCEval-v1](https://github.com/mtkresearch/MR-Models/tree/main/TC-Eval) and [ikala/tmmluplus](https://huggingface.co/datasets/ikala/tmmluplus). MMLU sources from [hails/mmlu_no_train](https://huggingface.co/datasets/hails/mmlu_no_train). We use the code revised from [EleutherAI/lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) to evaluate TMMLU+, DRCD, Table, and MMLU. \| Models \| \|↑ TMMLU+ (ACC) \| DRCD (EM) \| Table (ACC) \| MMLU (ACC) \| \|----------------------------------------------\|--------\|--------------\|-------------\|-------------\|------------\| \| \| \|TC, Knowledge \|TC, Reasoning\|TC, Reasoning\|EN, Knowledge\| \| \| \| 5 shot \| 3 shot \| 5 shot \| 5 shot \| \| [Yi-34B](https://huggingface.co/01-ai/Yi-34B)\| 34B \| 63.10 \| 84.57 \| 49.31 \| 77.42 \| \| [Qwen-14B](https://huggingface.co/01-ai/Qwen/Qwen-14B)\| 14B \| 51.30 \| 16.95 * \| 50.69 \| 68.83 \| \| [Yi-6B](https://huggingface.co/01-ai/Yi-6B) \| 6B \| 49.63 \| 76.61 \| 34.72 \| 65.35 \| \| [Qwen-7B](https://huggingface.co/01-ai/Qwen/Qwen-7B)\| 7B \| 42.84 \| 0.0 * \| 39.58 \| 61.00 \| \| [Breeze-7B-Base-v0.1](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v0.1) \| 7B \| 40.35 \| 81.13 \| 28.47 \| 61.63 \| \| [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1)\| 7B \| 36.93 \| 79.27 \| 27.78 \| 64.89 \| \* Few-shot learning cannot effectively guide the model to generate the proper answer. ## Chat Model Performance TMMLU+, DRCD, Table, and MT-Bench-tw source from [MediaTek-Research/TCEval-v2](https://huggingface.co/datasets/MediaTek-Research/TCEval-v2). [MediaTek-Research/TCEval-v2](https://huggingface.co/datasets/MediaTek-Research/TCEval-v2) derives from [TCEval-v1](https://github.com/mtkresearch/MR-Models/tree/main/TC-Eval) and [ikala/tmmluplus](https://huggingface.co/datasets/ikala/tmmluplus). MMLU sources from [hails/mmlu_no_train](https://huggingface.co/datasets/hails/mmlu_no_train). MT-Bench source from [lmsys/mt_bench_human_judgments](https://huggingface.co/datasets/lmsys/mt_bench_human_judgments). We use the code revised from [EleutherAI/lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) to evaluate TMMLU+, DRCD, Table, and MMLU. We use the code revised from [fastchat llm_judge](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) (GPT4 as judge) to evaluate MT-Bench-tw and MT-Bench. \| Models \| \|↑ MT-Bench-tw (Score)\| TMMLU+ (ACC) \| TMMLU+ (ACC) \| DRCD (EM) \| Table (ACC) \| MT-Bench (Score) \| MMLU (ACC) \| MMLU (ACC) \| \|---------------------------------------------------------------------------------------------------------\|--------\|--------------------\|--------------\|--------------\|-------------\|-------------\|------------------\|-------------\|-------------\| \| \| \|TC, Chat \|TC, Knowledge \|TC, Knowledge \|TC, Reasoning\|TC, Reasoning\|EN, Chat \|EN, Knowledge\|EN, Knowledge\| \| \| \|0 shot \| 0 shot \| 5 shot \| 3 shot \| 0 shot \|0 shot \| 0 shot \| 5 shot \| \| [gpt-3.5-turbo](https://openai.com) \| \|7.1 \| 41.76 \| \| \| \|7.9 \| 70.00 \| \| \| [Yi-34B-Chat](https://huggingface.co/01-ai/Yi-34B-Chat) \| 34B \|6.9 \| 54.87 \| \| \| 36.81 \|7.6 \| 71.04 \| \| \| [Qwen-14B-Chat](https://huggingface.co/Qwen/Qwen-14B-Chat) \| 14B \|6.4 \| 48.41 \| \| \| 41.67 \|7.2 \| 64.91 \| \| \| [Breeze-7B-Instruct-v0.1](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-v0.1) \| 7B \|5.7 \| 41.61 \| \| \| 45.83 \|7.1 \| 63.26 \| \| \| [Breeze-7B-Instruct-64k-v0.1](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-64k-v0.1) \| 7B \|5.5 \| 40.99 \| \| \| 36.11 \|7.1 \| 63.68 \| \| \| [Qwen-7B-Chat](https://huggingface.co/Qwen/Qwen-7B-Chat) \| 7B \|5.4 \| 40.02 \| \| \| 33.33 \|6.2 \| 55.94 \| \| \| [Yi-6B-Chat](https://huggingface.co/01-ai/Yi-6B-Chat) \| 6B \|5.0 \| 44.79 \| \| \| 25.69 \|6.0 \| 59.45 \| \| \| [Taiwan-LLM-13B-v2.0-chat](https://huggingface.co/yentinglin/Taiwan-LLM-13B-v2.0-chat) \| 13B \|5.0 \| 29.47 \| \| \| 23.61 \|-* \| 50.50 \| \| \| [Taiwan-LLM-7B-v2.1-chat](https://huggingface.co/yentinglin/Taiwan-LLM-7B-v2.1-chat) \| 7B \|4.2 \| 28.08 \| \| \| 31.25 \| -* \| 42.72 \| \| \* Taiwan-LLM models responds to multi-turn questions (English) in Traditional Chinese. Category Score of MT-Bench-tw (0 shot) \| Models \| STEM \|Extraction\|Reasoning\| Math \| Coding \| Roleplay\| Writing \|Humanities\|↑ AVG \| \|-----------------------------------------------------\|---------\|---------\|---------\|---------\|---------\|---------\|---------\|---------\|---------\| \| gpt-3.5-turbo \| 7.8 \| 6.1 \| 5.1 \| 6.4 \| 6.2 \| 8.7 \| 7.4 \| 9.3 \| 7.1 \| \| Yi-34B-Chat \| 9.0 \| 4.8 \| 5.7 \| 4.0 \| 4.7 \| 8.5 \| 8.7 \| 9.8 \| 6.9 \| \| Qwen-14B-Chat \| 7.6 \| 5.7 \| 4.5 \| 4.2 \| 5.3 \| 7.5 \| 7.3 \| 9.1 \| 6.4 \| \| Breeze-7B-Instruct-v0.1 \| 6.5 \| 5.6 \| 3.9 \| 3.6 \| 4.3 \| 6.9 \| 5.7 \| 9.3 \| 5.7 \| \| Breeze-7B-Instruct-64k-v0.1 \| 6.1 \| 5.3 \| 3.7 \| 2.9 \| 4.2 \| 7.0 \| 6.7 \| 8.3 \| 5.5 \| \| Qwen-7B-Chat \| 6.6 \| 4.5 \| 4.8 \| 2.9 \| 3.6 \| 6.2 \| 6.8 \| 8.2 \| 5.4 \| \| Yi-6B-Chat \| 7.3 \| 2.7 \| 3.1 \| 3.3 \| 2.3 \| 7.2 \| 5.2 \| 8.8 \| 5.0 \| \| Taiwan-LLM-13B-v2.0-chat \| 6.1 \| 3.4 \| 4.1 \| 2.3 \| 3.1 \| 7.4 \| 6.6 \| 6.8 \| 5.0 \| \| Taiwan-LLM-7B-v2.1-chat \| 5.2 \| 2.6 \| 2.3 \| 1.2 \| 3.4 \| 6.6 \| 5.7 \| 6.8 \| 4.2 \| Category ACC of TMMLU+ (0 shot) \| Model \| STEM \| Social Science \| Humanities \| Other \| ↑ AVG \| \|-----------------------------------------------------\|--------------\|----------------\|------------\|------------\|---------\| \| Yi-34B-Chat \| 47.65 \| 64.25 \| 52.73 \| 54.91 \| 54.87 \| \| Qwen-14B-Chat \| 43.83 \| 55.00 \| 48.55 \| 46.22 \| 48.41 \| \| Yi-6B-Chat \| 37.80 \| 51.74 \| 45.36 \| 44.25 \| 44.79 \| \| gpt-3.5-turbo \| 41.56 \| 46.72 \| 36.73 \| 42.03 \| 41.76 \| \| Breeze-7B-Instruct-v0.1 \| 37.41 \| 46.81 \| 42.06 \| 40.16 \| 41.61 \| \| Breeze-7B-Instruct-64k-v0.1 \| 37.88 \| 46.35 \| 40.31 \| 39.40 \| 40.99 \| \| Qwen-7B-Chat \| 35.44 \| 46.22 \| 38.35 \| 40.06 \| 40.02 \| \| Taiwan-LLM-13B-v2.0-chat \| 27.74 \| 33.69 \| 27.03 \| 29.43 \| 29.47 \| \| Taiwan-LLM-7B-v2.1-chat \| 25.58 \| 31.76 \| 27.36 \| 27.61 \| 28.08 \| ## Inference Performance In this test, we use the first 700 characters of the [web article](https://health.udn.com/health/story/5976/7699252?from=udn_ch1005_main_index) as the input and ask the model to write the same article again. All inferences run on 2 RTX A6000 GPUs (using `vllm`, with a tensor-parallel size of 2). \| Models \| ↓ Inference Time (sec)\|Estimated Max Input Length (Char)\| \|--------------------------------------------------------------------\|-------------------\|--------------------------\| \| Yi-6B \| 10.62 \| 5.2k \| \| Breeze-7B-Instruct-v0.1 \| 10.74 \| 11.1k \| \| Breeze-7B-Instruct-64k-v0.1 \| 10.74 \| 88.8k \| \| Qwen-7B \| 10.86 \| 9.8k \| \| Qwen-14B \| 18.89 \| 9.8k \| \| Mistral-7B-v0.1 \| 20.48 \| 5.1k \| \| Taiwan-LLM-7B-v2.1-base \| 26.26 \| 2.2k \| \| Taiwan-LLM-13B-v2.0-base \| 36.80 \| 2.2k \| \| Yi-34B \| 43.71 \| 4.5k \| ## Long-context Performance TBD ## Examples TBD ## Use in Transformers First install direct dependencies: ``` pip install transformers torch accelerate ``` If you want faster inference using flash-attention2, you need to install these dependencies: ```bash pip install packaging ninja pip install flash-attn ``` Then load the model in transformers: ```python from transformers import AutoModelForCausalLM, AutoTokenizer import torch model = AutoModelForCausalLM.from_pretrained( model="MediaTek-Research/Breeze-7B-Instruct-v0.1", device_map="auto", torch_dtype=torch.bfloat16, use_flash_attn_2=True # optional ) ``` The structure of the query template follows that of Mistral-7B-Instruct, as shown below. ```txt <s> SYS_PROMPT [INST] QUERY1 [/INST] RESPONSE1 [INST] QUERY2 [/INST] ``` where `SYS_PROMPT`, `QUERY1`, `RESPONSE1`, and `QUERY2` can be provided by the user. The suggested default `SYS_PROMPT` is ```txt You are a helpful AI assistant built by MediaTek Research. The user you are helping speaks Traditional Chinese and comes from Taiwan. ``` ## Citation ``` @article{breeze7b2024, title={}, author={}, journal={arXiv}, year={2024} } ```
Jongdae/koalpaca-polyglot-12.8b-bill	Jongdae	2024-01-13T00:41:05Z	0	0	peft	[ "peft", "safetensors", "arxiv:1910.09700", "base_model:beomi/polyglot-ko-12.8b-safetensors", "base_model:adapter:beomi/polyglot-ko-12.8b-safetensors", "region:us" ]	null	2024-01-13T00:41:02Z	--- library_name: peft base_model: beomi/polyglot-ko-12.8b-safetensors --- # Model Card for Model ID <!-- Provide a quick summary of what the model is/does. --> ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> - Developed by: [More Information Needed] - Funded by [optional]: [More Information Needed] - Shared by [optional]: [More Information Needed] - Model type: [More Information Needed] - Language(s) (NLP): [More Information Needed] - License: [More Information Needed] - Finetuned from model [optional]: [More Information Needed] ### Model Sources [optional] <!-- Provide the basic links for the model. --> - Repository: [More Information Needed] - Paper [optional]: [More Information Needed] - Demo [optional]: [More Information Needed] ## Uses <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> [More Information Needed] ### Downstream Use [optional] <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app --> [More Information Needed] ### Out-of-Scope Use <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> [More Information Needed] ## Bias, Risks, and Limitations <!-- This section is meant to convey both technical and sociotechnical limitations. --> [More Information Needed] ### Recommendations <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. --> Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. ## How to Get Started with the Model Use the code below to get started with the model. [More Information Needed] ## Training Details ### Training Data <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> [More Information Needed] ### Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> #### Preprocessing [optional] [More Information Needed] #### Training Hyperparameters - Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision --> #### Speeds, Sizes, Times [optional] <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. --> [More Information Needed] ## Evaluation <!-- This section describes the evaluation protocols and provides the results. --> ### Testing Data, Factors & Metrics #### Testing Data <!-- This should link to a Dataset Card if possible. --> [More Information Needed] #### Factors <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. --> [More Information Needed] #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> [More Information Needed] ### Results [More Information Needed] #### Summary ## Model Examination [optional] <!-- Relevant interpretability work for the model goes here --> [More Information Needed] ## Environmental Impact <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly --> Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - Hardware Type: [More Information Needed] - Hours used: [More Information Needed] - Cloud Provider: [More Information Needed] - Compute Region: [More Information Needed] - Carbon Emitted: [More Information Needed] ## Technical Specifications [optional] ### Model Architecture and Objective [More Information Needed] ### Compute Infrastructure [More Information Needed] #### Hardware [More Information Needed] #### Software [More Information Needed] ## Citation [optional] <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> BibTeX: [More Information Needed] APA: [More Information Needed] ## Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> [More Information Needed] ## More Information [optional] [More Information Needed] ## Model Card Authors [optional] [More Information Needed] ## Model Card Contact [More Information Needed] ### Framework versions - PEFT 0.7.2.dev0
NimaYaqmuri/speecht5_tts_voxpopuli_nl	NimaYaqmuri	2024-01-13T00:39:58Z	62	0	transformers	[ "transformers", "tensorboard", "safetensors", "speecht5", "text-to-audio", "Nima", "generated_from_trainer", "nl", "dataset:facebook/voxpopuli", "base_model:microsoft/speecht5_tts", "base_model:finetune:microsoft/speecht5_tts", "license:mit", "endpoints_compatible", "region:us" ]	text-to-audio	2024-01-12T12:29:58Z	--- language: - nl license: mit base_model: microsoft/speecht5_tts tags: - Nima - generated_from_trainer datasets: - facebook/voxpopuli model-index: - name: SpeechT5 TTS Dutch results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # SpeechT5 TTS Dutch This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the VoxPopuli dataset. It achieves the following results on the evaluation set: - Loss: 0.4569 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 1e-05 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 2 - total_train_batch_size: 32 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - training_steps: 4000 - mixed_precision_training: Native AMP ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| 0.5189 \| 4.3 \| 1000 \| 0.4776 \| \| 0.4994 \| 8.6 \| 2000 \| 0.4643 \| \| 0.4979 \| 12.9 \| 3000 \| 0.4589 \| \| 0.496 \| 17.2 \| 4000 \| 0.4569 \| ### Framework versions - Transformers 4.37.0.dev0 - Pytorch 2.1.2+cu121 - Datasets 2.4.0 - Tokenizers 0.15.0
ntc-ai/SDXL-LoRA-slider.the-starry-night	ntc-ai	2024-01-13T00:21:29Z	13	0	diffusers	[ "diffusers", "text-to-image", "stable-diffusion-xl", "lora", "template:sd-lora", "template:sdxl-lora", "sdxl-sliders", "ntcai.xyz-sliders", "concept", "en", "base_model:stabilityai/stable-diffusion-xl-base-1.0", "base_model:adapter:stabilityai/stable-diffusion-xl-base-1.0", "license:mit", "region:us" ]	text-to-image	2024-01-13T00:21:26Z	--- language: - en thumbnail: "images/evaluate/the starry night.../the starry night_17_3.0.png" widget: - text: the starry night output: url: images/the starry night_17_3.0.png - text: the starry night output: url: images/the starry night_19_3.0.png - text: the starry night output: url: images/the starry night_20_3.0.png - text: the starry night output: url: images/the starry night_21_3.0.png - text: the starry night output: url: images/the starry night_22_3.0.png tags: - text-to-image - stable-diffusion-xl - lora - template:sd-lora - template:sdxl-lora - sdxl-sliders - ntcai.xyz-sliders - concept - diffusers license: "mit" inference: false instance_prompt: "the starry night" base_model: "stabilityai/stable-diffusion-xl-base-1.0" --- # ntcai.xyz slider - the starry night (SDXL LoRA) \| Strength: -3 \| Strength: 0 \| Strength: 3 \| \| --- \| --- \| --- \| \| <img src="images/the starry night_17_-3.0.png" width=256 height=256 /> \| <img src="images/the starry night_17_0.0.png" width=256 height=256 /> \| <img src="images/the starry night_17_3.0.png" width=256 height=256 /> \| \| <img src="images/the starry night_19_-3.0.png" width=256 height=256 /> \| <img src="images/the starry night_19_0.0.png" width=256 height=256 /> \| <img src="images/the starry night_19_3.0.png" width=256 height=256 /> \| \| <img src="images/the starry night_20_-3.0.png" width=256 height=256 /> \| <img src="images/the starry night_20_0.0.png" width=256 height=256 /> \| <img src="images/the starry night_20_3.0.png" width=256 height=256 /> \| ## Download Weights for this model are available in Safetensors format. ## Trigger words You can apply this LoRA with trigger words for additional effect: ``` the starry night ``` ## Use in diffusers ```python from diffusers import StableDiffusionXLPipeline from diffusers import EulerAncestralDiscreteScheduler import torch pipe = StableDiffusionXLPipeline.from_single_file("https://huggingface.co/martyn/sdxl-turbo-mario-merge-top-rated/blob/main/topRatedTurboxlLCM_v10.safetensors") pipe.to("cuda") pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config) # Load the LoRA pipe.load_lora_weights('ntc-ai/SDXL-LoRA-slider.the-starry-night', weight_name='the starry night.safetensors', adapter_name="the starry night") # Activate the LoRA pipe.set_adapters(["the starry night"], adapter_weights=[2.0]) prompt = "medieval rich kingpin sitting in a tavern, the starry night" negative_prompt = "nsfw" width = 512 height = 512 num_inference_steps = 10 guidance_scale = 2 image = pipe(prompt, negative_prompt=negative_prompt, width=width, height=height, guidance_scale=guidance_scale, num_inference_steps=num_inference_steps).images[0] image.save('result.png') ``` ## Support the Patreon If you like this model please consider [joining our Patreon](https://www.patreon.com/NTCAI). By joining our Patreon, you'll gain access to an ever-growing library of over 1070+ unique and diverse LoRAs, covering a wide range of styles and genres. You'll also receive early access to new models and updates, exclusive behind-the-scenes content, and the powerful LoRA slider creator, allowing you to craft your own custom LoRAs and experiment with endless possibilities. Your support on Patreon will allow us to continue developing and refining new models. ## Other resources - [CivitAI](https://civitai.com/user/ntc) - Follow ntc on Civit for even more LoRAs - [ntcai.xyz](https://ntcai.xyz) - See ntcai.xyz to find more articles and LoRAs
cnatale/Mistral-7B-Instruct-v0_1-Txt-2-Presto-SQL	cnatale	2024-01-13T00:14:40Z	5	0	peft	[ "peft", "tensorboard", "safetensors", "trl", "sft", "generated_from_trainer", "dataset:generator", "base_model:mistralai/Mistral-7B-Instruct-v0.1", "base_model:adapter:mistralai/Mistral-7B-Instruct-v0.1", "license:apache-2.0", "region:us" ]	null	2024-01-12T18:27:30Z	--- license: apache-2.0 library_name: peft tags: - trl - sft - generated_from_trainer datasets: - generator base_model: mistralai/Mistral-7B-Instruct-v0.1 model-index: - name: Mistral-7B-Instruct-v0_1-Txt-2-Presto-SQL results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # Mistral-7B-Instruct-v0_1-Txt-2-Presto-SQL This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the generator dataset. It achieves the following results on the evaluation set: - Loss: 0.6454 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001 - train_batch_size: 4 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: constant - lr_scheduler_warmup_steps: 0.03 - training_steps: 80 - mixed_precision_training: Native AMP ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| 1.3516 \| 0.71 \| 10 \| 1.0778 \| \| 1.0131 \| 1.43 \| 20 \| 0.8707 \| \| 0.8446 \| 2.14 \| 30 \| 0.7695 \| \| 0.7563 \| 2.86 \| 40 \| 0.7202 \| \| 0.7009 \| 3.57 \| 50 \| 0.6803 \| \| 0.6368 \| 4.29 \| 60 \| 0.6585 \| \| 0.6201 \| 5.0 \| 70 \| 0.6473 \| \| 0.5755 \| 5.71 \| 80 \| 0.6454 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.1.2+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
dsteiner93/dqn-SpaceInvadersNoFrameskip-v4	dsteiner93	2024-01-13T00:13:50Z	3	0	stable-baselines3	[ "stable-baselines3", "SpaceInvadersNoFrameskip-v4", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]	reinforcement-learning	2024-01-13T00:13:19Z	--- library_name: stable-baselines3 tags: - SpaceInvadersNoFrameskip-v4 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: DQN results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: SpaceInvadersNoFrameskip-v4 type: SpaceInvadersNoFrameskip-v4 metrics: - type: mean_reward value: 459.50 +/- 101.94 name: mean_reward verified: false --- # DQN Agent playing SpaceInvadersNoFrameskip-v4 This is a trained model of a DQN agent playing SpaceInvadersNoFrameskip-v4 using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3) and the [RL Zoo](https://github.com/DLR-RM/rl-baselines3-zoo). The RL Zoo is a training framework for Stable Baselines3 reinforcement learning agents, with hyperparameter optimization and pre-trained agents included. ## Usage (with SB3 RL Zoo) RL Zoo: https://github.com/DLR-RM/rl-baselines3-zoo<br/> SB3: https://github.com/DLR-RM/stable-baselines3<br/> SB3 Contrib: https://github.com/Stable-Baselines-Team/stable-baselines3-contrib Install the RL Zoo (with SB3 and SB3-Contrib): ```bash pip install rl_zoo3 ``` ``` # Download model and save it into the logs/ folder python -m rl_zoo3.load_from_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -orga dsteiner93 -f logs/ python -m rl_zoo3.enjoy --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ ``` If you installed the RL Zoo3 via pip (`pip install rl_zoo3`), from anywhere you can do: ``` python -m rl_zoo3.load_from_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -orga dsteiner93 -f logs/ python -m rl_zoo3.enjoy --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ ``` ## Training (with the RL Zoo) ``` python -m rl_zoo3.train --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ # Upload the model and generate video (when possible) python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ -orga dsteiner93 ``` ## Hyperparameters ```python OrderedDict([('batch_size', 32), ('buffer_size', 110000), ('env_wrapper', ['stable_baselines3.common.atari_wrappers.AtariWrapper']), ('exploration_final_eps', 0.01), ('exploration_fraction', 0.1), ('frame_stack', 4), ('gradient_steps', 1), ('learning_rate', 'linear_1e-4'), ('learning_starts', 100000), ('n_timesteps', 1000000), ('optimize_memory_usage', False), ('policy', 'CnnPolicy'), ('target_update_interval', 1000), ('train_freq', 4), ('normalize', False)]) ``` # Environment Arguments ```python {'render_mode': 'rgb_array'} ```
TheBloke/TowerInstruct-7B-v0.1-GPTQ	TheBloke	2024-01-13T00:10:41Z	14	2	transformers	[ "transformers", "safetensors", "llama", "text-generation", "translation", "en", "de", "fr", "zh", "pt", "nl", "ru", "ko", "it", "es", "base_model:Unbabel/TowerInstruct-7B-v0.1", "base_model:quantized:Unbabel/TowerInstruct-7B-v0.1", "license:cc-by-nc-4.0", "autotrain_compatible", "text-generation-inference", "4-bit", "gptq", "region:us" ]	translation	2024-01-12T23:50:11Z	--- base_model: Unbabel/TowerInstruct-7B-v0.1 inference: false language: - en - de - fr - zh - pt - nl - ru - ko - it - es license: cc-by-nc-4.0 metrics: - comet model_creator: Unbabel model_name: TowerInstruct 7B v0.1 model_type: llama pipeline_tag: translation prompt_template: '<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ' quantized_by: TheBloke --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # TowerInstruct 7B v0.1 - GPTQ - Model creator: [Unbabel](https://huggingface.co/Unbabel) - Original model: [TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1) <!-- description start --> # Description This repo contains GPTQ model files for [Unbabel's TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1). Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them. <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF) * [Unbabel's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: ChatML ``` <\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ``` <!-- prompt-template end --> <!-- licensing start --> ## Licensing The creator of the source model has listed its license as `cc-by-nc-4.0`, and this quantization has therefore used that same license. As this model is based on Llama 2, it is also subject to the Meta Llama 2 license terms, and the license files for that are additionally included. It should therefore be considered as being claimed to be licensed under both licenses. I contacted Hugging Face for clarification on dual licensing but they do not yet have an official position. Should this change, or should Meta provide any feedback on this situation, I will update this section accordingly. In the meantime, any questions regarding licensing, and in particular how these two licenses might interact, should be directed to the original model repository: [Unbabel's TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1). <!-- licensing end --> <!-- README_GPTQ.md-compatible clients start --> ## Known compatible clients / servers GPTQ models are currently supported on Linux (NVidia/AMD) and Windows (NVidia only). macOS users: please use GGUF models. These GPTQ models are known to work in the following inference servers/webuis. - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) - [KoboldAI United](https://github.com/henk717/koboldai) - [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui) - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) This may not be a complete list; if you know of others, please let me know! <!-- README_GPTQ.md-compatible clients end --> <!-- README_GPTQ.md-provided-files start --> ## Provided files, and GPTQ parameters Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements. Each separate quant is in a different branch. See below for instructions on fetching from different branches. Most GPTQ files are made with AutoGPTQ. Mistral models are currently made with Transformers. <details> <summary>Explanation of GPTQ parameters</summary> - Bits: The bit size of the quantised model. - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value. - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now. - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy. - GPTQ dataset: The calibration dataset used during quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ calibration dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s). - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences. - ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama and Mistral models in 4-bit. </details> \| Branch \| Bits \| GS \| Act Order \| Damp % \| GPTQ Dataset \| Seq Len \| Size \| ExLlama \| Desc \| \| ------ \| ---- \| -- \| --------- \| ------ \| ------------ \| ------- \| ---- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ/tree/main) \| 4 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 3.90 GB \| Yes \| 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. \| \| [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ/tree/gptq-4bit-32g-actorder_True) \| 4 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.28 GB \| Yes \| 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. \| \| [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ/tree/gptq-8bit--1g-actorder_True) \| 8 \| None \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 7.01 GB \| No \| 8-bit, with Act Order. No group size, to lower VRAM requirements. \| \| [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ/tree/gptq-8bit-128g-actorder_True) \| 8 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 7.16 GB \| No \| 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. \| \| [gptq-8bit-32g-actorder_True](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ/tree/gptq-8bit-32g-actorder_True) \| 8 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 7.62 GB \| No \| 8-bit, with group size 32g and Act Order for maximum inference quality. \| \| [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ/tree/gptq-4bit-64g-actorder_True) \| 4 \| 64 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.02 GB \| Yes \| 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. \| <!-- README_GPTQ.md-provided-files end --> <!-- README_GPTQ.md-download-from-branches start --> ## How to download, including from branches ### In text-generation-webui To download from the `main` branch, enter `TheBloke/TowerInstruct-7B-v0.1-GPTQ` in the "Download model" box. To download from another branch, add `:branchname` to the end of the download name, eg `TheBloke/TowerInstruct-7B-v0.1-GPTQ:gptq-4bit-32g-actorder_True` ### From the command line I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` To download the `main` branch to a folder called `TowerInstruct-7B-v0.1-GPTQ`: ```shell mkdir TowerInstruct-7B-v0.1-GPTQ huggingface-cli download TheBloke/TowerInstruct-7B-v0.1-GPTQ --local-dir TowerInstruct-7B-v0.1-GPTQ --local-dir-use-symlinks False ``` To download from a different branch, add the `--revision` parameter: ```shell mkdir TowerInstruct-7B-v0.1-GPTQ huggingface-cli download TheBloke/TowerInstruct-7B-v0.1-GPTQ --revision gptq-4bit-32g-actorder_True --local-dir TowerInstruct-7B-v0.1-GPTQ --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage</summary> If you remove the `--local-dir-use-symlinks False` parameter, the files will instead be stored in the central Hugging Face cache directory (default location on Linux is: `~/.cache/huggingface`), and symlinks will be added to the specified `--local-dir`, pointing to their real location in the cache. This allows for interrupted downloads to be resumed, and allows you to quickly clone the repo to multiple places on disk without triggering a download again. The downside, and the reason why I don't list that as the default option, is that the files are then hidden away in a cache folder and it's harder to know where your disk space is being used, and to clear it up if/when you want to remove a download model. The cache location can be changed with the `HF_HOME` environment variable, and/or the `--cache-dir` parameter to `huggingface-cli`. For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell mkdir TowerInstruct-7B-v0.1-GPTQ HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/TowerInstruct-7B-v0.1-GPTQ --local-dir TowerInstruct-7B-v0.1-GPTQ --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> ### With `git` (not recommended) To clone a specific branch with `git`, use a command like this: ```shell git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ ``` Note that using Git with HF repos is strongly discouraged. It will be much slower than using `huggingface-hub`, and will use twice as much disk space as it has to store the model files twice (it stores every byte both in the intended target folder, and again in the `.git` folder as a blob.) <!-- README_GPTQ.md-download-from-branches end --> <!-- README_GPTQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/TowerInstruct-7B-v0.1-GPTQ`. - To download from a specific branch, enter for example `TheBloke/TowerInstruct-7B-v0.1-GPTQ:gptq-4bit-32g-actorder_True` - see Provided Files above for the list of branches for each option. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `TowerInstruct-7B-v0.1-GPTQ` 7. The model will automatically load, and is now ready for use! 8. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. - Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file `quantize_config.json`. 9. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_GPTQ.md-text-generation-webui end --> <!-- README_GPTQ.md-use-from-tgi start --> ## Serving this model from Text Generation Inference (TGI) It's recommended to use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/TowerInstruct-7B-v0.1-GPTQ --port 3000 --quantize gptq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires huggingface-hub 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' client = InferenceClient(endpoint_url) response = client.text_generation( prompt_template, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(f"Model output: {response}") ``` <!-- README_GPTQ.md-use-from-tgi end --> <!-- README_GPTQ.md-use-from-python start --> ## Python code example: inference from this GPTQ model ### Install the necessary packages Requires: Transformers 4.33.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later. ```shell pip3 install --upgrade transformers optimum # If using PyTorch 2.1 + CUDA 12.x: pip3 install --upgrade auto-gptq # or, if using PyTorch 2.1 + CUDA 11.x: pip3 install --upgrade auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ ``` If you are using PyTorch 2.0, you will need to install AutoGPTQ from source. Likewise if you have problems with the pre-built wheels, you should try building from source: ```shell pip3 uninstall -y auto-gptq git clone https://github.com/PanQiWei/AutoGPTQ cd AutoGPTQ git checkout v0.5.1 pip3 install . ``` ### Example Python code ```python from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline model_name_or_path = "TheBloke/TowerInstruct-7B-v0.1-GPTQ" # To use a different branch, change revision # For example: revision="gptq-4bit-32g-actorder_True" model = AutoModelForCausalLM.from_pretrained(model_name_or_path, device_map="auto", trust_remote_code=False, revision="main") tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True) prompt = "Write a story about llamas" system_message = "You are a story writing assistant" prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' print("\n\n* Generate:") input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda() output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512) print(tokenizer.decode(output[0])) # Inference can also be done using transformers' pipeline print("* Pipeline:") pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(pipe(prompt_template)[0]['generated_text']) ``` <!-- README_GPTQ.md-use-from-python end --> <!-- README_GPTQ.md-compatibility start --> ## Compatibility The files provided are tested to work with Transformers. For non-Mistral models, AutoGPTQ can also be used directly. [ExLlama](https://github.com/turboderp/exllama) is compatible with Llama architecture models (including Mistral, Yi, DeepSeek, SOLAR, etc) in 4-bit. Please see the Provided Files table above for per-file compatibility. For a list of clients/servers, please see "Known compatible clients / servers", above. <!-- README_GPTQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Unbabel's TowerInstruct 7B v0.1 # Model Card for TowerInstruct-7B-v0.1 ## Model Details ### Model Description TowerInstruct-7B is a language model that results from fine-tuning TowerBase on the TowerBlocks supervised fine-tuning dataset. TowerInstruct-7B-v0.1 is the first model in the series. The model is trained to handle several translation-related tasks, such as general machine translation (e.g., sentence- and document-level translation, terminology-aware translation, context-aware translation), automatic post edition, named-entity recognition, gramatical error correction, and paraphrase generation. We will release more details in the upcoming technical report. - Developed by: Unbabel, Instituto Superior Técnico, CentraleSupélec University of Paris-Saclay - Model type: A 7B parameter model fine-tuned on a mix of publicly available, synthetic datasets on translation-related tasks, as well as conversational datasets and code instructions. - Language(s) (NLP): English, Portuguese, Spanish, French, German, Dutch, Italian, Korean, Chinese, Russian - License: CC-BY-NC-4.0, Llama 2 is licensed under the [LLAMA 2 Community License](https://ai.meta.com/llama/license/), Copyright © Meta Platforms, Inc. All Rights Reserved. - Finetuned from model: [TowerBase](https://huggingface.co/Unbabel/TowerBase-7B-v0.1) ## Intended uses & limitations The model was initially fine-tuned on a filtered and preprocessed supervised fine-tuning dataset ([TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1)), which contains a diverse range of data sources: - Translation - Automatic Post Edition - Machine Translation Evaluation - Context-aware Translation - Terminology-aware Translation - Multi-reference Translation - Named-entity Recognition - Paraphrase Generation - Synthetic Chat data - Code instructions You can find the dataset and all data sources of [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1) here. Here's how you can run the model using the `pipeline()` function from 🤗 Transformers: ```python # Install transformers from source - only needed for versions <= v4.34 # pip install git+https://github.com/huggingface/transformers.git # pip install accelerate import torch from transformers import pipeline pipe = pipeline("text-generation", model="Unbabel/TowerInstruct-v0.1", torch_dtype=torch.bfloat16, device_map="auto") # We use the tokenizer’s chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating messages = [ {"role": "user", "content": "Translate the following text from Portuguese into English.\nPortuguese: Um grupo de investigadores lançou um novo modelo para tarefas relacionadas com tradução.\nEnglish:"}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=256, do_sample=False) print(outputs[0]["generated_text"]) # <\|im_start\|>user # Translate the following text from Portuguese into English. # Portuguese: Um grupo de investigadores lançou um novo modelo para tarefas relacionadas com tradução. # English:<\|im_end\|> # <\|im_start\|>assistant # A group of researchers has launched a new model for translation-related tasks. ``` ### Out-of-Scope Use The model is not guaranteed to perform for languages other than the 10 languages it supports. Even though we trained the model on conversational data and code instructions, it is not intended to be used as a conversational chatbot or code assistant. ## Bias, Risks, and Limitations TowerInstruct-v0.1 has not been aligned to human preferences, so the model may generate problematic outputs (e.g., hallucinations, harmful content, or false statements). ## Prompt Format TowerInstruct-v0.1 was trained using the ChatML prompt templates without any system prompts. An example follows below: ``` <\|im_start\|>user {USER PROMPT}<\|im_end\|> <\|im_start\|>assistant {MODEL RESPONSE}<\|im_end\|> <\|im_start\|>user [...] ``` ### Supervised tasks The prompts for all supervised tasks can be found in [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1). We have used multiple prompt templates for each task. While different prompts may offer different outputs, the difference in downstream performance should be very minimal. ## Training Details ### Training Data Link to [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1). #### Training Hyperparameters The following hyperparameters were used during training: - total_train_batch_size: 256 - learning_rate: 7e-06 - lr_scheduler_type: cosine - lr_scheduler_warmup_steps: 500 - weight_decay: 0.01 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - num_epochs: 4 - max_seq_length: 2048 ## Citation To be completed. [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
TheBloke/TowerInstruct-7B-v0.1-AWQ	TheBloke	2024-01-13T00:01:55Z	45	1	transformers	[ "transformers", "safetensors", "llama", "text-generation", "translation", "en", "de", "fr", "zh", "pt", "nl", "ru", "ko", "it", "es", "base_model:Unbabel/TowerInstruct-7B-v0.1", "base_model:quantized:Unbabel/TowerInstruct-7B-v0.1", "license:cc-by-nc-4.0", "autotrain_compatible", "text-generation-inference", "4-bit", "awq", "region:us" ]	translation	2024-01-12T23:50:11Z	--- base_model: Unbabel/TowerInstruct-7B-v0.1 inference: false language: - en - de - fr - zh - pt - nl - ru - ko - it - es license: cc-by-nc-4.0 metrics: - comet model_creator: Unbabel model_name: TowerInstruct 7B v0.1 model_type: llama pipeline_tag: translation prompt_template: '<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ' quantized_by: TheBloke --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # TowerInstruct 7B v0.1 - AWQ - Model creator: [Unbabel](https://huggingface.co/Unbabel) - Original model: [TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1) <!-- description start --> ## Description This repo contains AWQ model files for [Unbabel's TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1). ### About AWQ AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference with equivalent or better quality compared to the most commonly used GPTQ settings. AWQ models are currently supported on Linux and Windows, with NVidia GPUs only. macOS users: please use GGUF models instead. It is supported by: - [Text Generation Webui](https://github.com/oobabooga/text-generation-webui) - using Loader: AutoAWQ - [vLLM](https://github.com/vllm-project/vllm) - version 0.2.2 or later for support for all model types. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later, from any code or client that supports Transformers - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) - for use from Python code <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF) * [Unbabel's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: ChatML ``` <\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ``` <!-- prompt-template end --> <!-- licensing start --> ## Licensing The creator of the source model has listed its license as `cc-by-nc-4.0`, and this quantization has therefore used that same license. As this model is based on Llama 2, it is also subject to the Meta Llama 2 license terms, and the license files for that are additionally included. It should therefore be considered as being claimed to be licensed under both licenses. I contacted Hugging Face for clarification on dual licensing but they do not yet have an official position. Should this change, or should Meta provide any feedback on this situation, I will update this section accordingly. In the meantime, any questions regarding licensing, and in particular how these two licenses might interact, should be directed to the original model repository: [Unbabel's TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1). <!-- licensing end --> <!-- README_AWQ.md-provided-files start --> ## Provided files, and AWQ parameters I currently release 128g GEMM models only. The addition of group_size 32 models, and GEMV kernel models, is being actively considered. Models are released as sharded safetensors files. \| Branch \| Bits \| GS \| AWQ Dataset \| Seq Len \| Size \| \| ------ \| ---- \| -- \| ----------- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-AWQ/tree/main) \| 4 \| 128 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 3.89 GB <!-- README_AWQ.md-provided-files end --> <!-- README_AWQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/TowerInstruct-7B-v0.1-AWQ`. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `TowerInstruct-7B-v0.1-AWQ` 7. Select Loader: AutoAWQ. 8. Click Load, and the model will load and is now ready for use. 9. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. 10. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_AWQ.md-text-generation-webui end --> <!-- README_AWQ.md-use-from-vllm start --> ## Multi-user inference server: vLLM Documentation on installing and using vLLM [can be found here](https://vllm.readthedocs.io/en/latest/). - Please ensure you are using vLLM version 0.2 or later. - When using vLLM as a server, pass the `--quantization awq` parameter. For example: ```shell python3 -m vllm.entrypoints.api_server --model TheBloke/TowerInstruct-7B-v0.1-AWQ --quantization awq --dtype auto ``` - When using vLLM from Python code, again set `quantization=awq`. For example: ```python from vllm import LLM, SamplingParams prompts = [ "Tell me about AI", "Write a story about llamas", "What is 291 - 150?", "How much wood would a woodchuck chuck if a woodchuck could chuck wood?", ] prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' prompts = [prompt_template.format(prompt=prompt) for prompt in prompts] sampling_params = SamplingParams(temperature=0.8, top_p=0.95) llm = LLM(model="TheBloke/TowerInstruct-7B-v0.1-AWQ", quantization="awq", dtype="auto") outputs = llm.generate(prompts, sampling_params) # Print the outputs. for output in outputs: prompt = output.prompt generated_text = output.outputs[0].text print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}") ``` <!-- README_AWQ.md-use-from-vllm start --> <!-- README_AWQ.md-use-from-tgi start --> ## Multi-user inference server: Hugging Face Text Generation Inference (TGI) Use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/TowerInstruct-7B-v0.1-AWQ --port 3000 --quantize awq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires [huggingface-hub](https://github.com/huggingface/huggingface_hub) 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' client = InferenceClient(endpoint_url) response = client.text_generation(prompt, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1) print(f"Model output: ", response) ``` <!-- README_AWQ.md-use-from-tgi end --> <!-- README_AWQ.md-use-from-python start --> ## Inference from Python code using Transformers ### Install the necessary packages - Requires: [Transformers](https://huggingface.co/docs/transformers) 4.35.0 or later. - Requires: [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) 0.1.6 or later. ```shell pip3 install --upgrade "autoawq>=0.1.6" "transformers>=4.35.0" ``` Note that if you are using PyTorch 2.0.1, the above AutoAWQ command will automatically upgrade you to PyTorch 2.1.0. If you are using CUDA 11.8 and wish to continue using PyTorch 2.0.1, instead run this command: ```shell pip3 install https://github.com/casper-hansen/AutoAWQ/releases/download/v0.1.6/autoawq-0.1.6+cu118-cp310-cp310-linux_x86_64.whl ``` If you have problems installing [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) using the pre-built wheels, install it from source instead: ```shell pip3 uninstall -y autoawq git clone https://github.com/casper-hansen/AutoAWQ cd AutoAWQ pip3 install . ``` ### Transformers example code (requires Transformers 4.35.0 and later) ```python from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer model_name_or_path = "TheBloke/TowerInstruct-7B-v0.1-AWQ" tokenizer = AutoTokenizer.from_pretrained(model_name_or_path) model = AutoModelForCausalLM.from_pretrained( model_name_or_path, low_cpu_mem_usage=True, device_map="cuda:0" ) # Using the text streamer to stream output one token at a time streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True) prompt = "Tell me about AI" prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' # Convert prompt to tokens tokens = tokenizer( prompt_template, return_tensors='pt' ).input_ids.cuda() generation_params = { "do_sample": True, "temperature": 0.7, "top_p": 0.95, "top_k": 40, "max_new_tokens": 512, "repetition_penalty": 1.1 } # Generate streamed output, visible one token at a time generation_output = model.generate( tokens, streamer=streamer, generation_params ) # Generation without a streamer, which will include the prompt in the output generation_output = model.generate( tokens, generation_params ) # Get the tokens from the output, decode them, print them token_output = generation_output[0] text_output = tokenizer.decode(token_output) print("model.generate output: ", text_output) # Inference is also possible via Transformers' pipeline from transformers import pipeline pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, *generation_params ) pipe_output = pipe(prompt_template)[0]['generated_text'] print("pipeline output: ", pipe_output) ``` <!-- README_AWQ.md-use-from-python end --> <!-- README_AWQ.md-compatibility start --> ## Compatibility The files provided are tested to work with: - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) using `Loader: AutoAWQ`. - [vLLM](https://github.com/vllm-project/vllm) version 0.2.0 and later. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) version 1.1.0 and later. - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later. - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) version 0.1.1 and later. <!-- README_AWQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Unbabel's TowerInstruct 7B v0.1 # Model Card for TowerInstruct-7B-v0.1 ## Model Details ### Model Description TowerInstruct-7B is a language model that results from fine-tuning TowerBase on the TowerBlocks supervised fine-tuning dataset. TowerInstruct-7B-v0.1 is the first model in the series. The model is trained to handle several translation-related tasks, such as general machine translation (e.g., sentence- and document-level translation, terminology-aware translation, context-aware translation), automatic post edition, named-entity recognition, gramatical error correction, and paraphrase generation. We will release more details in the upcoming technical report. - Developed by: Unbabel, Instituto Superior Técnico, CentraleSupélec University of Paris-Saclay - Model type: A 7B parameter model fine-tuned on a mix of publicly available, synthetic datasets on translation-related tasks, as well as conversational datasets and code instructions. - Language(s) (NLP): English, Portuguese, Spanish, French, German, Dutch, Italian, Korean, Chinese, Russian - License: CC-BY-NC-4.0, Llama 2 is licensed under the [LLAMA 2 Community License](https://ai.meta.com/llama/license/), Copyright © Meta Platforms, Inc. All Rights Reserved. - Finetuned from model: [TowerBase](https://huggingface.co/Unbabel/TowerBase-7B-v0.1) ## Intended uses & limitations The model was initially fine-tuned on a filtered and preprocessed supervised fine-tuning dataset ([TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1)), which contains a diverse range of data sources: - Translation - Automatic Post Edition - Machine Translation Evaluation - Context-aware Translation - Terminology-aware Translation - Multi-reference Translation - Named-entity Recognition - Paraphrase Generation - Synthetic Chat data - Code instructions You can find the dataset and all data sources of [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1) here. Here's how you can run the model using the `pipeline()` function from 🤗 Transformers: ```python # Install transformers from source - only needed for versions <= v4.34 # pip install git+https://github.com/huggingface/transformers.git # pip install accelerate import torch from transformers import pipeline pipe = pipeline("text-generation", model="Unbabel/TowerInstruct-v0.1", torch_dtype=torch.bfloat16, device_map="auto") # We use the tokenizer’s chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating messages = [ {"role": "user", "content": "Translate the following text from Portuguese into English.\nPortuguese: Um grupo de investigadores lançou um novo modelo para tarefas relacionadas com tradução.\nEnglish:"}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=256, do_sample=False) print(outputs[0]["generated_text"]) # <\|im_start\|>user # Translate the following text from Portuguese into English. # Portuguese: Um grupo de investigadores lançou um novo modelo para tarefas relacionadas com tradução. # English:<\|im_end\|> # <\|im_start\|>assistant # A group of researchers has launched a new model for translation-related tasks. ``` ### Out-of-Scope Use The model is not guaranteed to perform for languages other than the 10 languages it supports. Even though we trained the model on conversational data and code instructions, it is not intended to be used as a conversational chatbot or code assistant. ## Bias, Risks, and Limitations TowerInstruct-v0.1 has not been aligned to human preferences, so the model may generate problematic outputs (e.g., hallucinations, harmful content, or false statements). ## Prompt Format TowerInstruct-v0.1 was trained using the ChatML prompt templates without any system prompts. An example follows below: ``` <\|im_start\|>user {USER PROMPT}<\|im_end\|> <\|im_start\|>assistant {MODEL RESPONSE}<\|im_end\|> <\|im_start\|>user [...] ``` ### Supervised tasks The prompts for all supervised tasks can be found in [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1). We have used multiple prompt templates for each task. While different prompts may offer different outputs, the difference in downstream performance should be very minimal. ## Training Details ### Training Data Link to [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1). #### Training Hyperparameters The following hyperparameters were used during training: - total_train_batch_size: 256 - learning_rate: 7e-06 - lr_scheduler_type: cosine - lr_scheduler_warmup_steps: 500 - weight_decay: 0.01 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - num_epochs: 4 - max_seq_length: 2048 ## Citation To be completed. [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
TheBloke/TowerInstruct-7B-v0.1-GGUF	TheBloke	2024-01-12T23:57:32Z	215	16	transformers	[ "transformers", "gguf", "llama", "translation", "en", "de", "fr", "zh", "pt", "nl", "ru", "ko", "it", "es", "base_model:Unbabel/TowerInstruct-7B-v0.1", "base_model:quantized:Unbabel/TowerInstruct-7B-v0.1", "license:cc-by-nc-4.0", "region:us", "conversational" ]	translation	2024-01-12T23:50:11Z	--- base_model: Unbabel/TowerInstruct-7B-v0.1 inference: false language: - en - de - fr - zh - pt - nl - ru - ko - it - es license: cc-by-nc-4.0 metrics: - comet model_creator: Unbabel model_name: TowerInstruct 7B v0.1 model_type: llama pipeline_tag: translation prompt_template: '<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ' quantized_by: TheBloke --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # TowerInstruct 7B v0.1 - GGUF - Model creator: [Unbabel](https://huggingface.co/Unbabel) - Original model: [TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1) <!-- description start --> ## Description This repo contains GGUF format model files for [Unbabel's TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1). <!-- description end --> <!-- README_GGUF.md-about-gguf start --> ### About GGUF GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp. Here is an incomplete list of clients and libraries that are known to support GGUF: * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option. * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration. * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling. * [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel. * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023. * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection. * [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration. * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server. * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use. * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models. <!-- README_GGUF.md-about-gguf end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF) * [Unbabel's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: ChatML ``` <\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ``` <!-- prompt-template end --> <!-- licensing start --> ## Licensing The creator of the source model has listed its license as `cc-by-nc-4.0`, and this quantization has therefore used that same license. As this model is based on Llama 2, it is also subject to the Meta Llama 2 license terms, and the license files for that are additionally included. It should therefore be considered as being claimed to be licensed under both licenses. I contacted Hugging Face for clarification on dual licensing but they do not yet have an official position. Should this change, or should Meta provide any feedback on this situation, I will update this section accordingly. In the meantime, any questions regarding licensing, and in particular how these two licenses might interact, should be directed to the original model repository: [Unbabel's TowerInstruct 7B v0.1](https://huggingface.co/Unbabel/TowerInstruct-7B-v0.1). <!-- licensing end --> <!-- compatibility_gguf start --> ## Compatibility These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) They are also compatible with many third party UIs and libraries - please see the list at the top of this README. ## Explanation of quantisation methods <details> <summary>Click to see details</summary> The new methods available are: * GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw) * GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw. * GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw. * GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw * GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw Refer to the Provided Files table below to see what files use which methods, and how. </details> <!-- compatibility_gguf end --> <!-- README_GGUF.md-provided-files start --> ## Provided files \| Name \| Quant method \| Bits \| Size \| Max RAM required \| Use case \| \| ---- \| ---- \| ---- \| ---- \| ---- \| ----- \| \| [towerinstruct-7b-v0.1.Q2_K.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q2_K.gguf) \| Q2_K \| 2 \| 2.53 GB\| 5.03 GB \| smallest, significant quality loss - not recommended for most purposes \| \| [towerinstruct-7b-v0.1.Q3_K_S.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q3_K_S.gguf) \| Q3_K_S \| 3 \| 2.95 GB\| 5.45 GB \| very small, high quality loss \| \| [towerinstruct-7b-v0.1.Q3_K_M.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q3_K_M.gguf) \| Q3_K_M \| 3 \| 3.30 GB\| 5.80 GB \| very small, high quality loss \| \| [towerinstruct-7b-v0.1.Q3_K_L.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q3_K_L.gguf) \| Q3_K_L \| 3 \| 3.60 GB\| 6.10 GB \| small, substantial quality loss \| \| [towerinstruct-7b-v0.1.Q4_0.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q4_0.gguf) \| Q4_0 \| 4 \| 3.83 GB\| 6.33 GB \| legacy; small, very high quality loss - prefer using Q3_K_M \| \| [towerinstruct-7b-v0.1.Q4_K_S.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q4_K_S.gguf) \| Q4_K_S \| 4 \| 3.86 GB\| 6.36 GB \| small, greater quality loss \| \| [towerinstruct-7b-v0.1.Q4_K_M.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q4_K_M.gguf) \| Q4_K_M \| 4 \| 4.08 GB\| 6.58 GB \| medium, balanced quality - recommended \| \| [towerinstruct-7b-v0.1.Q5_0.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q5_0.gguf) \| Q5_0 \| 5 \| 4.65 GB\| 7.15 GB \| legacy; medium, balanced quality - prefer using Q4_K_M \| \| [towerinstruct-7b-v0.1.Q5_K_S.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q5_K_S.gguf) \| Q5_K_S \| 5 \| 4.65 GB\| 7.15 GB \| large, low quality loss - recommended \| \| [towerinstruct-7b-v0.1.Q5_K_M.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q5_K_M.gguf) \| Q5_K_M \| 5 \| 4.78 GB\| 7.28 GB \| large, very low quality loss - recommended \| \| [towerinstruct-7b-v0.1.Q6_K.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q6_K.gguf) \| Q6_K \| 6 \| 5.53 GB\| 8.03 GB \| very large, extremely low quality loss \| \| [towerinstruct-7b-v0.1.Q8_0.gguf](https://huggingface.co/TheBloke/TowerInstruct-7B-v0.1-GGUF/blob/main/towerinstruct-7b-v0.1.Q8_0.gguf) \| Q8_0 \| 8 \| 7.16 GB\| 9.66 GB \| very large, extremely low quality loss - not recommended \| Note: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead. <!-- README_GGUF.md-provided-files end --> <!-- README_GGUF.md-how-to-download start --> ## How to download GGUF files Note for manual downloaders: You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file. The following clients/libraries will automatically download models for you, providing a list of available models to choose from: * LM Studio * LoLLMS Web UI * Faraday.dev ### In `text-generation-webui` Under Download Model, you can enter the model repo: TheBloke/TowerInstruct-7B-v0.1-GGUF and below it, a specific filename to download, such as: towerinstruct-7b-v0.1.Q4_K_M.gguf. Then click Download. ### On the command line, including multiple files at once I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` Then you can download any individual model file to the current directory, at high speed, with a command like this: ```shell huggingface-cli download TheBloke/TowerInstruct-7B-v0.1-GGUF towerinstruct-7b-v0.1.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage (click to read)</summary> You can also download multiple files at once with a pattern: ```shell huggingface-cli download TheBloke/TowerInstruct-7B-v0.1-GGUF --local-dir . --local-dir-use-symlinks False --include='Q4_Kgguf' ``` For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/TowerInstruct-7B-v0.1-GGUF towerinstruct-7b-v0.1.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> <!-- README_GGUF.md-how-to-download end --> <!-- README_GGUF.md-how-to-run start --> ## Example `llama.cpp` command Make sure you are using `llama.cpp` from commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later. ```shell ./main -ngl 35 -m towerinstruct-7b-v0.1.Q4_K_M.gguf --color -c 4096 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "<\|im_start\|>system\n{system_message}<\|im_end\|>\n<\|im_start\|>user\n{prompt}<\|im_end\|>\n<\|im_start\|>assistant" ``` Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration. Change `-c 4096` to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically. Note that longer sequence lengths require much more resources, so you may need to reduce this value. If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins` For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md) ## How to run in `text-generation-webui` Further instructions can be found in the text-generation-webui documentation, here: [text-generation-webui/docs/04 ‐ Model Tab.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/04%20%E2%80%90%20Model%20Tab.md#llamacpp). ## How to run from Python code You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) or [ctransformers](https://github.com/marella/ctransformers) libraries. Note that at the time of writing (Nov 27th 2023), ctransformers has not been updated for some time and is not compatible with some recent models. Therefore I recommend you use llama-cpp-python. ### How to load this model in Python code, using llama-cpp-python For full documentation, please see: [llama-cpp-python docs](https://abetlen.github.io/llama-cpp-python/). #### First install the package Run one of the following commands, according to your system: ```shell # Base ctransformers with no GPU acceleration pip install llama-cpp-python # With NVidia CUDA acceleration CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install llama-cpp-python # Or with OpenBLAS acceleration CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python # Or with CLBLast acceleration CMAKE_ARGS="-DLLAMA_CLBLAST=on" pip install llama-cpp-python # Or with AMD ROCm GPU acceleration (Linux only) CMAKE_ARGS="-DLLAMA_HIPBLAS=on" pip install llama-cpp-python # Or with Metal GPU acceleration for macOS systems only CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python # In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA: $env:CMAKE_ARGS = "-DLLAMA_OPENBLAS=on" pip install llama-cpp-python ``` #### Simple llama-cpp-python example code ```python from llama_cpp import Llama # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system. llm = Llama( model_path="./towerinstruct-7b-v0.1.Q4_K_M.gguf", # Download the model file first n_ctx=4096, # The max sequence length to use - note that longer sequence lengths require much more resources n_threads=8, # The number of CPU threads to use, tailor to your system and the resulting performance n_gpu_layers=35 # The number of layers to offload to GPU, if you have GPU acceleration available ) # Simple inference example output = llm( "<\|im_start\|>system\n{system_message}<\|im_end\|>\n<\|im_start\|>user\n{prompt}<\|im_end\|>\n<\|im_start\|>assistant", # Prompt max_tokens=512, # Generate up to 512 tokens stop=["</s>"], # Example stop token - not necessarily correct for this specific model! Please check before using. echo=True # Whether to echo the prompt ) # Chat Completion API llm = Llama(model_path="./towerinstruct-7b-v0.1.Q4_K_M.gguf", chat_format="llama-2") # Set chat_format according to the model you are using llm.create_chat_completion( messages = [ {"role": "system", "content": "You are a story writing assistant."}, { "role": "user", "content": "Write a story about llamas." } ] ) ``` ## How to use with LangChain Here are guides on using llama-cpp-python and ctransformers with LangChain: * [LangChain + llama-cpp-python](https://python.langchain.com/docs/integrations/llms/llamacpp) * [LangChain + ctransformers](https://python.langchain.com/docs/integrations/providers/ctransformers) <!-- README_GGUF.md-how-to-run end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> <!-- original-model-card start --> # Original model card: Unbabel's TowerInstruct 7B v0.1 # Model Card for TowerInstruct-7B-v0.1 ## Model Details ### Model Description TowerInstruct-7B is a language model that results from fine-tuning TowerBase on the TowerBlocks supervised fine-tuning dataset. TowerInstruct-7B-v0.1 is the first model in the series. The model is trained to handle several translation-related tasks, such as general machine translation (e.g., sentence- and document-level translation, terminology-aware translation, context-aware translation), automatic post edition, named-entity recognition, gramatical error correction, and paraphrase generation. We will release more details in the upcoming technical report. - Developed by: Unbabel, Instituto Superior Técnico, CentraleSupélec University of Paris-Saclay - Model type: A 7B parameter model fine-tuned on a mix of publicly available, synthetic datasets on translation-related tasks, as well as conversational datasets and code instructions. - Language(s) (NLP): English, Portuguese, Spanish, French, German, Dutch, Italian, Korean, Chinese, Russian - License: CC-BY-NC-4.0, Llama 2 is licensed under the [LLAMA 2 Community License](https://ai.meta.com/llama/license/), Copyright © Meta Platforms, Inc. All Rights Reserved. - Finetuned from model: [TowerBase](https://huggingface.co/Unbabel/TowerBase-7B-v0.1) ## Intended uses & limitations The model was initially fine-tuned on a filtered and preprocessed supervised fine-tuning dataset ([TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1)), which contains a diverse range of data sources: - Translation - Automatic Post Edition - Machine Translation Evaluation - Context-aware Translation - Terminology-aware Translation - Multi-reference Translation - Named-entity Recognition - Paraphrase Generation - Synthetic Chat data - Code instructions You can find the dataset and all data sources of [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1) here. Here's how you can run the model using the `pipeline()` function from 🤗 Transformers: ```python # Install transformers from source - only needed for versions <= v4.34 # pip install git+https://github.com/huggingface/transformers.git # pip install accelerate import torch from transformers import pipeline pipe = pipeline("text-generation", model="Unbabel/TowerInstruct-v0.1", torch_dtype=torch.bfloat16, device_map="auto") # We use the tokenizer’s chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating messages = [ {"role": "user", "content": "Translate the following text from Portuguese into English.\nPortuguese: Um grupo de investigadores lançou um novo modelo para tarefas relacionadas com tradução.\nEnglish:"}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=256, do_sample=False) print(outputs[0]["generated_text"]) # <\|im_start\|>user # Translate the following text from Portuguese into English. # Portuguese: Um grupo de investigadores lançou um novo modelo para tarefas relacionadas com tradução. # English:<\|im_end\|> # <\|im_start\|>assistant # A group of researchers has launched a new model for translation-related tasks. ``` ### Out-of-Scope Use The model is not guaranteed to perform for languages other than the 10 languages it supports. Even though we trained the model on conversational data and code instructions, it is not intended to be used as a conversational chatbot or code assistant. ## Bias, Risks, and Limitations TowerInstruct-v0.1 has not been aligned to human preferences, so the model may generate problematic outputs (e.g., hallucinations, harmful content, or false statements). ## Prompt Format TowerInstruct-v0.1 was trained using the ChatML prompt templates without any system prompts. An example follows below: ``` <\|im_start\|>user {USER PROMPT}<\|im_end\|> <\|im_start\|>assistant {MODEL RESPONSE}<\|im_end\|> <\|im_start\|>user [...] ``` ### Supervised tasks The prompts for all supervised tasks can be found in [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1). We have used multiple prompt templates for each task. While different prompts may offer different outputs, the difference in downstream performance should be very minimal. ## Training Details ### Training Data Link to [TowerBlocks](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.1). #### Training Hyperparameters The following hyperparameters were used during training: - total_train_batch_size: 256 - learning_rate: 7e-06 - lr_scheduler_type: cosine - lr_scheduler_warmup_steps: 500 - weight_decay: 0.01 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - num_epochs: 4 - max_seq_length: 2048 ## Citation To be completed. [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl) <!-- original-model-card end -->
chinmaydk99/messi_LoRA	chinmaydk99	2024-01-12T23:53:37Z	1	0	diffusers	[ "diffusers", "tensorboard", "stable-diffusion-xl", "stable-diffusion-xl-diffusers", "text-to-image", "lora", "template:sd-lora", "base_model:stabilityai/stable-diffusion-xl-base-1.0", "base_model:adapter:stabilityai/stable-diffusion-xl-base-1.0", "license:openrail++", "region:us" ]	text-to-image	2024-01-12T23:53:08Z	--- tags: - stable-diffusion-xl - stable-diffusion-xl-diffusers - text-to-image - diffusers - lora - template:sd-lora base_model: stabilityai/stable-diffusion-xl-base-1.0 instance_prompt: a photo of TOK messi license: openrail++ --- # SDXL LoRA DreamBooth - chinmaydk99/messi_LoRA <Gallery /> ## Model description These are chinmaydk99/messi_LoRA LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0. The weights were trained using [DreamBooth](https://dreambooth.github.io/). LoRA for the text encoder was enabled: False. Special VAE used for training: madebyollin/sdxl-vae-fp16-fix. ## Trigger words You should use a photo of TOK messi to trigger the image generation. ## Download model Weights for this model are available in Safetensors format. [Download](chinmaydk99/messi_LoRA/tree/main) them in the Files & versions tab.
TheBloke/PiVoT-SUS-RP-GPTQ	TheBloke	2024-01-12T23:49:50Z	8	1	transformers	[ "transformers", "safetensors", "llama", "text-generation", "base_model:maywell/PiVoT-SUS-RP", "base_model:quantized:maywell/PiVoT-SUS-RP", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "4-bit", "gptq", "region:us" ]	text-generation	2024-01-12T18:57:15Z	--- base_model: maywell/PiVoT-SUS-RP inference: false license: apache-2.0 model_creator: Jeonghwan Park model_name: Pivot SUS RP model_type: yi prompt_template: '{prompt} ' quantized_by: TheBloke --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Pivot SUS RP - GPTQ - Model creator: [Jeonghwan Park](https://huggingface.co/maywell) - Original model: [Pivot SUS RP](https://huggingface.co/maywell/PiVoT-SUS-RP) <!-- description start --> # Description This repo contains GPTQ model files for [Jeonghwan Park's Pivot SUS RP](https://huggingface.co/maywell/PiVoT-SUS-RP). Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them. <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/PiVoT-SUS-RP-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GGUF) * [Jeonghwan Park's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/maywell/PiVoT-SUS-RP) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: Unknown ``` {prompt} ``` <!-- prompt-template end --> <!-- README_GPTQ.md-compatible clients start --> ## Known compatible clients / servers GPTQ models are currently supported on Linux (NVidia/AMD) and Windows (NVidia only). macOS users: please use GGUF models. These GPTQ models are known to work in the following inference servers/webuis. - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) - [KoboldAI United](https://github.com/henk717/koboldai) - [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui) - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) This may not be a complete list; if you know of others, please let me know! <!-- README_GPTQ.md-compatible clients end --> <!-- README_GPTQ.md-provided-files start --> ## Provided files, and GPTQ parameters Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements. Each separate quant is in a different branch. See below for instructions on fetching from different branches. Most GPTQ files are made with AutoGPTQ. Mistral models are currently made with Transformers. <details> <summary>Explanation of GPTQ parameters</summary> - Bits: The bit size of the quantised model. - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value. - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now. - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy. - GPTQ dataset: The calibration dataset used during quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ calibration dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s). - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences. - ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama and Mistral models in 4-bit. </details> \| Branch \| Bits \| GS \| Act Order \| Damp % \| GPTQ Dataset \| Seq Len \| Size \| ExLlama \| Desc \| \| ------ \| ---- \| -- \| --------- \| ------ \| ------------ \| ------- \| ---- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ/tree/main) \| 4 \| None \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 18.60 GB \| Yes \| 4-bit, with Act Order. No group size, to lower VRAM requirements. \| \| [gptq-4bit-128g-actorder_True](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ/tree/gptq-4bit-128g-actorder_True) \| 4 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 19.25 GB \| Yes \| 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. \| \| [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ/tree/gptq-4bit-32g-actorder_True) \| 4 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 21.21 GB \| Yes \| 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. \| \| [gptq-3bit-128g-actorder_True](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ/tree/gptq-3bit-128g-actorder_True) \| 3 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 15.03 GB \| No \| 3-bit, with group size 128g and act-order. Higher quality than 128g-False. \| \| [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ/tree/gptq-8bit--1g-actorder_True) \| 8 \| None \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 35.34 GB \| No \| 8-bit, with Act Order. No group size, to lower VRAM requirements. \| \| [gptq-3bit-32g-actorder_True](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ/tree/gptq-3bit-32g-actorder_True) \| 3 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 16.90 GB \| No \| 3-bit, with group size 64g and act-order. Highest quality 3-bit option. \| \| [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ/tree/gptq-8bit-128g-actorder_True) \| 8 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 36.11 GB \| No \| 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. \| <!-- README_GPTQ.md-provided-files end --> <!-- README_GPTQ.md-download-from-branches start --> ## How to download, including from branches ### In text-generation-webui To download from the `main` branch, enter `TheBloke/PiVoT-SUS-RP-GPTQ` in the "Download model" box. To download from another branch, add `:branchname` to the end of the download name, eg `TheBloke/PiVoT-SUS-RP-GPTQ:gptq-4bit-128g-actorder_True` ### From the command line I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` To download the `main` branch to a folder called `PiVoT-SUS-RP-GPTQ`: ```shell mkdir PiVoT-SUS-RP-GPTQ huggingface-cli download TheBloke/PiVoT-SUS-RP-GPTQ --local-dir PiVoT-SUS-RP-GPTQ --local-dir-use-symlinks False ``` To download from a different branch, add the `--revision` parameter: ```shell mkdir PiVoT-SUS-RP-GPTQ huggingface-cli download TheBloke/PiVoT-SUS-RP-GPTQ --revision gptq-4bit-128g-actorder_True --local-dir PiVoT-SUS-RP-GPTQ --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage</summary> If you remove the `--local-dir-use-symlinks False` parameter, the files will instead be stored in the central Hugging Face cache directory (default location on Linux is: `~/.cache/huggingface`), and symlinks will be added to the specified `--local-dir`, pointing to their real location in the cache. This allows for interrupted downloads to be resumed, and allows you to quickly clone the repo to multiple places on disk without triggering a download again. The downside, and the reason why I don't list that as the default option, is that the files are then hidden away in a cache folder and it's harder to know where your disk space is being used, and to clear it up if/when you want to remove a download model. The cache location can be changed with the `HF_HOME` environment variable, and/or the `--cache-dir` parameter to `huggingface-cli`. For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell mkdir PiVoT-SUS-RP-GPTQ HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/PiVoT-SUS-RP-GPTQ --local-dir PiVoT-SUS-RP-GPTQ --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> ### With `git` (not recommended) To clone a specific branch with `git`, use a command like this: ```shell git clone --single-branch --branch gptq-4bit-128g-actorder_True https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ ``` Note that using Git with HF repos is strongly discouraged. It will be much slower than using `huggingface-hub`, and will use twice as much disk space as it has to store the model files twice (it stores every byte both in the intended target folder, and again in the `.git` folder as a blob.) <!-- README_GPTQ.md-download-from-branches end --> <!-- README_GPTQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/PiVoT-SUS-RP-GPTQ`. - To download from a specific branch, enter for example `TheBloke/PiVoT-SUS-RP-GPTQ:gptq-4bit-128g-actorder_True` - see Provided Files above for the list of branches for each option. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `PiVoT-SUS-RP-GPTQ` 7. The model will automatically load, and is now ready for use! 8. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. - Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file `quantize_config.json`. 9. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_GPTQ.md-text-generation-webui end --> <!-- README_GPTQ.md-use-from-tgi start --> ## Serving this model from Text Generation Inference (TGI) It's recommended to use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/PiVoT-SUS-RP-GPTQ --port 3000 --quantize gptq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires huggingface-hub 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''{prompt} ''' client = InferenceClient(endpoint_url) response = client.text_generation( prompt_template, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(f"Model output: {response}") ``` <!-- README_GPTQ.md-use-from-tgi end --> <!-- README_GPTQ.md-use-from-python start --> ## Python code example: inference from this GPTQ model ### Install the necessary packages Requires: Transformers 4.33.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later. ```shell pip3 install --upgrade transformers optimum # If using PyTorch 2.1 + CUDA 12.x: pip3 install --upgrade auto-gptq # or, if using PyTorch 2.1 + CUDA 11.x: pip3 install --upgrade auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ ``` If you are using PyTorch 2.0, you will need to install AutoGPTQ from source. Likewise if you have problems with the pre-built wheels, you should try building from source: ```shell pip3 uninstall -y auto-gptq git clone https://github.com/PanQiWei/AutoGPTQ cd AutoGPTQ git checkout v0.5.1 pip3 install . ``` ### Example Python code ```python from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline model_name_or_path = "TheBloke/PiVoT-SUS-RP-GPTQ" # To use a different branch, change revision # For example: revision="gptq-4bit-128g-actorder_True" model = AutoModelForCausalLM.from_pretrained(model_name_or_path, device_map="auto", trust_remote_code=False, revision="main") tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True) prompt = "Write a story about llamas" system_message = "You are a story writing assistant" prompt_template=f'''{prompt} ''' print("\n\n* Generate:") input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda() output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512) print(tokenizer.decode(output[0])) # Inference can also be done using transformers' pipeline print("* Pipeline:") pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(pipe(prompt_template)[0]['generated_text']) ``` <!-- README_GPTQ.md-use-from-python end --> <!-- README_GPTQ.md-compatibility start --> ## Compatibility The files provided are tested to work with Transformers. For non-Mistral models, AutoGPTQ can also be used directly. [ExLlama](https://github.com/turboderp/exllama) is compatible with Llama architecture models (including Mistral, Yi, DeepSeek, SOLAR, etc) in 4-bit. Please see the Provided Files table above for per-file compatibility. For a list of clients/servers, please see "Known compatible clients / servers", above. <!-- README_GPTQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Jeonghwan Park's Pivot SUS RP
LoneStriker/Bagel-Hermes-2x34b-5.0bpw-h6-exl2	LoneStriker	2024-01-12T23:49:33Z	8	1	transformers	[ "transformers", "safetensors", "mixtral", "text-generation", "yi", "moe", "conversational", "license:other", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T23:33:59Z	--- license: other license_name: yi-license license_link: https://huggingface.co/01-ai/Yi-34B-200K/blob/main/LICENSE tags: - yi - moe --- ![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/6468ce47e134d050a58aa89c/mLH2E0dk9On_LcFX9yhuS.jpeg) # Bagel-Hermes-2x34B This is the model for Bagel-Hermes-2x34B. I used [mergekit](https://github.com/cg123/mergekit) to make this MOE model. # Prompt Template(s): Since [bagel-dpo-34b-v0.2](https://huggingface.co/jondurbin/bagel-dpo-34b-v0.2) uses many prompt templates, and [Nous-Hermes-2-Yi-34B](https://huggingface.co/NousResearch/Nous-Hermes-2-Yi-34B) uses ChatML, you can utilize ChatML and other prompt templates provided by bagel. Note: I currently do not know which prompt template is best. ### ChatML: ``` <\|im_start\|>system {system}<\|im_end\|> <\|im_start\|>user {user}<\|im_end\|> <\|im_start\|>assistant {asistant}<\|im_end\|> ``` ### Alpaca (sort of) ``` Below is an instruction that describes a task. Write a response that appropriately completes the request. ### Instruction: {system} {instruction} ### Response: ``` ### Vicuna ``` {system} USER: {instruction} ASSISTANT: ``` Visit [bagel-dpo-34b-v0.2](https://huggingface.co/jondurbin/bagel-dpo-34b-v0.2) to try more prompt templates. # Yaml Config to reproduce ```yaml base_model: nontoxic-bagel-34b-v0.2 gate_mode: hidden dtype: bfloat16 experts: - source_model: bagel-dpo-34b-v0.2 positive_prompts: ["question answering", "Q:", science", "biology", "chemistry", "physics"] - source_model: Nous-Hermes-2-Yi-34B positive_prompts: ["chat", "math", "reason", "mathematics", "solve", "count", "python", "javascript", "programming", "algorithm", "tell me", "assistant"] ``` # Quantizationed versions Quantizationed versions of this model is available thanks to [TheBloke](https://hf.co/TheBloke). ##### GPTQ - [TheBloke/Bagel-Hermes-2x34B-GPTQ](https://huggingface.co/TheBloke/Bagel-Hermes-2x34B-GPTQ) ##### GGUF - [TheBloke/Bagel-Hermes-2x34B-GGUF](https://huggingface.co/TheBloke/Bagel-Hermes-2x34B-GGUF) ##### AWQ - [TheBloke/Bagel-Hermes-2x34B-AWQ](https://huggingface.co/TheBloke/Bagel-Hermes-2x34B-AWQ) If you would like to support me: [☕ Buy Me a Coffee](https://www.buymeacoffee.com/weyaxi)
jysssacc/opt-350m_adalora_lr0.05_bs10_epoch5_wd0.01	jysssacc	2024-01-12T23:45:25Z	0	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:facebook/opt-350m", "base_model:adapter:facebook/opt-350m", "license:other", "region:us" ]	null	2024-01-12T23:43:47Z	--- license: other library_name: peft tags: - generated_from_trainer base_model: facebook/opt-350m model-index: - name: opt-350m_adalora_lr0.05_bs10_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # opt-350m_adalora_lr0.05_bs10_epoch5_wd0.01 This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 9.2670 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.05 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| No log \| 1.0 \| 63 \| 3.8373 \| \| 4.0958 \| 2.0 \| 126 \| 5.9224 \| \| 4.0958 \| 3.0 \| 189 \| 8.2854 \| \| 6.616 \| 4.0 \| 252 \| 9.5332 \| \| 12.126 \| 5.0 \| 315 \| 9.2670 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
adamo1139/Yi-34B-200K-rawrr1-LORA-DPO-experimental-r2	adamo1139	2024-01-12T23:41:19Z	5	1	peft	[ "peft", "llama", "generated_from_trainer", "4-bit", "bitsandbytes", "region:us" ]	null	2024-01-12T23:10:48Z	--- library_name: peft tags: - generated_from_trainer model-index: - name: qlora-yi-34b-200k-rawrr-2 results: [] --- [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl) <details><summary>See axolotl config</summary> axolotl version: `0.3.0` ```yaml base_model: ./yi-34b-200k-llamafied base_model_config: ./yi-34b-200k-llamafied model_type: LlamaForCausalLM tokenizer_type: LlamaTokenizer is_mistral_derived_model: false is_llama_derived_model: true load_in_8bit: false load_in_4bit: true bnb_config_kwargs: llm_int8_has_fp16_weight: false bnb_4bit_quant_type: nf4 bnb_4bit_use_double_quant: true bnb_4bit_compute_dtype: torch.bfloat16 torch_dtype: bf16 strict: false rl: true datasets: - path: /run/media/.../axolotl/datasets/rawrr_v1/ split: train type: apply_chatml dataset_prepared_path: last_run_prepared val_set_size: 0.01 adapter: qlora lora_model_dir: sequence_len: 200 sample_packing: false lora_r: 4 lora_alpha: 8 lora_dropout: 0.05 lora_target_modules: - q_proj - v_proj - k_proj - o_proj - gate_proj - down_proj - up_proj lora_fan_in_fan_out: wandb_project: wandb_watch: wandb_run_id: wandb_log_model: output_dir: ./qlora-yi-34b-200k-rawrr-2 pad_to_sequence_len: true micro_batch_size: 1 gradient_accumulation_steps: 16 num_epochs: 1 optimizer: adamw_bnb_8bit torchdistx_path: lr_scheduler: cosine learning_rate: 0.00003 cosine_min_lr_ratio: 0.2 train_on_inputs: false group_by_length: false bf16: true fp16: false tf32: false bfloat16: true flash_optimum: false gradient_checkpointing: true early_stopping_patience: save_safetensors: local_rank: logging_steps: 1 xformers_attention: flash_attention: true deepspeed: seed: 42 warmup_steps: 100 eval_steps: 5000000 save_steps: 80 save_total_limit: 10 eval_table_size: eval_table_max_new_tokens: debug: weight_decay: fsdp: fsdp_config: special_tokens: bos_token: "<\|startoftext\|>" eos_token: "<\|endoftext\|>" unk_token: "<unk>" ``` </details><br> # qlora-yi-34b-200k-rawrr-2 Yi-34B-200K trained via DPO on rawrr_v1 dataset. Sequence length of just 200, that's the max I could fit in with axolotl on RTX 3090 Ti. ## Model description Looks like DPO worked, even despite being a tiny rank, sequence length and learning rate. \ That's great, I was preparing to start fine-tuning in the cloud but I saw [this](https://github.com/OpenAccess-AI-Collective/axolotl/pull/1060) pull request and I was hopeful it would allow me to squeeze in 34B qlora DPO on 24GB of VRAM. And it did! ## Intended uses & limitations merge it with yi-34b-200k llama-fied to get a more raw-feeling Yi 34B. In my initial testing on comparison between exl2 4.65bpw base and my DPO fine-tune, my model has no AALM refusals and feels moreso like a true base model. \ Some instruct training still shines through though when you ask it to make an itinerary etc for you. ## Training procedure The following `bitsandbytes` quantization config was used during training: - quant_method: bitsandbytes - load_in_8bit: False - load_in_4bit: True - llm_int8_threshold: 6.0 - llm_int8_skip_modules: None - llm_int8_enable_fp32_cpu_offload: False - llm_int8_has_fp16_weight: False - bnb_4bit_quant_type: nf4 - bnb_4bit_use_double_quant: True - bnb_4bit_compute_dtype: bfloat16 ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 3e-05 - train_batch_size: 1 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 16 - total_train_batch_size: 16 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 100 - training_steps: 517 ### Training results ### Framework versions - PEFT 0.7.0 - Transformers 4.37.0.dev0 - Pytorch 2.0.1+cu118 - Datasets 2.15.0 - Tokenizers 0.15.0
nech06/T5base_billsum_10000_1024_256	nech06	2024-01-12T23:13:45Z	90	0	transformers	[ "transformers", "tensorboard", "safetensors", "t5", "text2text-generation", "generated_from_trainer", "base_model:google-t5/t5-base", "base_model:finetune:google-t5/t5-base", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text2text-generation	2024-01-12T23:13:10Z	--- license: apache-2.0 base_model: t5-base tags: - generated_from_trainer metrics: - rouge model-index: - name: T5base_billsum_10000_1024_256 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # T5base_billsum_10000_1024_256 This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.8577 - Rouge1: 18.5154 - Rouge2: 14.7876 - Rougel: 17.9792 - Rougelsum: 18.0210 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 1 - eval_batch_size: 1 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3 ### Training results ### Framework versions - Transformers 4.35.2 - Pytorch 2.1.0+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
Utshav/bert-fined-tuned-cola	Utshav	2024-01-12T23:09:10Z	95	1	transformers	[ "transformers", "tensorboard", "safetensors", "bert", "text-classification", "generated_from_trainer", "base_model:google-bert/bert-base-cased", "base_model:finetune:google-bert/bert-base-cased", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2024-01-12T22:57:04Z	--- license: apache-2.0 base_model: bert-base-cased tags: - generated_from_trainer metrics: - matthews_correlation model-index: - name: bert-fined-tuned-cola results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # bert-fined-tuned-cola This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.7846 - Matthews Correlation: 0.5626 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Matthews Correlation \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:--------------------:\| \| 0.453 \| 1.0 \| 1069 \| 0.4373 \| 0.5189 \| \| 0.3124 \| 2.0 \| 2138 \| 0.6774 \| 0.5727 \| \| 0.1966 \| 3.0 \| 3207 \| 0.7846 \| 0.5626 \| ### Framework versions - Transformers 4.37.0.dev0 - Pytorch 2.1.0+cu121 - Tokenizers 0.15.0
Ricardo54321/TaxiV3	Ricardo54321	2024-01-12T23:08:31Z	0	0	null	[ "Taxi-v3", "q-learning", "reinforcement-learning", "custom-implementation", "model-index", "region:us" ]	reinforcement-learning	2024-01-12T22:39:38Z	--- tags: - Taxi-v3 - q-learning - reinforcement-learning - custom-implementation model-index: - name: TaxiV3 results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: Taxi-v3 type: Taxi-v3 metrics: - type: mean_reward value: 7.56 +/- 2.71 name: mean_reward verified: false --- # Q-Learning Agent playing1 Taxi-v3 This is a trained model of a Q-Learning agent playing Taxi-v3 . ## Usage ```python model = load_from_hub(repo_id="Ricardo54321/TaxiV3", filename="q-learning.pkl") # Don't forget to check if you need to add additional attributes (is_slippery=False etc) env = gym.make(model["env_id"]) ```
jysssacc/opt-350m_lora_lr0.05_bs10_epoch5_wd0.01	jysssacc	2024-01-12T23:05:17Z	0	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:facebook/opt-350m", "base_model:adapter:facebook/opt-350m", "license:other", "region:us" ]	null	2024-01-12T23:04:40Z	--- license: other library_name: peft tags: - generated_from_trainer base_model: facebook/opt-350m model-index: - name: opt-350m_lora_lr0.05_bs10_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # opt-350m_lora_lr0.05_bs10_epoch5_wd0.01 This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 7.9621 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.05 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| No log \| 1.0 \| 63 \| 4.7369 \| \| 5.5016 \| 2.0 \| 126 \| 7.4478 \| \| 5.5016 \| 3.0 \| 189 \| 7.4983 \| \| 7.317 \| 4.0 \| 252 \| 7.8890 \| \| 7.2033 \| 5.0 \| 315 \| 7.9621 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
QuangDuy/Bloom-3b	QuangDuy	2024-01-12T23:00:54Z	4	0	peft	[ "peft", "tensorboard", "safetensors", "trl", "sft", "generated_from_trainer", "base_model:bigscience/bloom-3b", "base_model:adapter:bigscience/bloom-3b", "license:bigscience-bloom-rail-1.0", "region:us" ]	null	2024-01-12T15:01:34Z	--- license: bigscience-bloom-rail-1.0 library_name: peft tags: - trl - sft - generated_from_trainer base_model: bigscience/bloom-3b model-index: - name: Bloom-3b-v3 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # Bloom-3b-v3 This model is a fine-tuned version of [bigscience/bloom-3b](https://huggingface.co/bigscience/bloom-3b) on an unknown dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0002 - train_batch_size: 2 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 8 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine - lr_scheduler_warmup_ratio: 0.05 - num_epochs: 2 - mixed_precision_training: Native AMP ### Training results ### Framework versions - PEFT 0.7.2.dev0 - Transformers 4.37.0.dev0 - Pytorch 2.0.0 - Datasets 2.16.1 - Tokenizers 0.15.0
andrewma5/FoodBase-distilBERT	andrewma5	2024-01-12T22:56:48Z	46	0	transformers	[ "transformers", "tf", "distilbert", "token-classification", "generated_from_keras_callback", "base_model:distilbert/distilbert-base-uncased", "base_model:finetune:distilbert/distilbert-base-uncased", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	token-classification	2024-01-12T22:48:31Z	--- license: apache-2.0 base_model: distilbert-base-uncased tags: - generated_from_keras_callback model-index: - name: andrewma5/FoodBase-distilBERT results: [] --- <!-- This model card has been generated automatically according to the information Keras had access to. You should probably proofread and complete it, then remove this comment. --> # andrewma5/FoodBase-distilBERT This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset. It achieves the following results on the evaluation set: - Train Loss: 0.0368 - Validation Loss: 0.0443 - Epoch: 4 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - optimizer: {'inner_optimizer': {'module': 'transformers.optimization_tf', 'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 190, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'decay': 0.0, 'beta_1': 0.8999999761581421, 'beta_2': 0.9990000128746033, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}, 'registered_name': 'AdamWeightDecay'}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000} - training_precision: mixed_float16 ### Training results \| Train Loss \| Validation Loss \| Epoch \| \|:----------:\|:---------------:\|:-----:\| \| 0.5478 \| 0.2080 \| 0 \| \| 0.1247 \| 0.0643 \| 1 \| \| 0.0532 \| 0.0492 \| 2 \| \| 0.0416 \| 0.0447 \| 3 \| \| 0.0368 \| 0.0443 \| 4 \| ### Framework versions - Transformers 4.35.2 - TensorFlow 2.15.0 - Datasets 2.16.1 - Tokenizers 0.15.0
neuralmagic/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K-quant-ds	neuralmagic	2024-01-12T22:47:29Z	0	8	null	[ "onnx", "deepsparse", "zero-shot-classification", "base_model:laion/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K", "base_model:quantized:laion/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K", "region:us" ]	zero-shot-classification	2023-12-19T18:03:47Z	--- pipeline_tag: zero-shot-classification base_model: laion/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K inference: false tags: - deepsparse --- This is a [SparseML](https://github.com/neuralmagic/sparseml) quantized version of https://huggingface.co/laion/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K that is ready to use with the [DeepSparse](https://github.com/neuralmagic/deepsparse) CPU inference engine. It achieves 71.1% zero-shot top-1 accuracy on ImageNet and 95.6% zero-shot top-1 accuracy on Imagenette. For comparison the dense version (the original model) achieves 72.8% on ImageNet and 95.7% on Imagenette. On an Intel avx512 CPU machine with 64 cores and VNNI support, this model achieves a 2.35x speedup for textual and 2.84x speedup for visual inputs as compared to the full-precision model. With a batch size of 64, the throughput was measured as 1230 items/sec for images and 2009 items/sec for text. Notebook for basic usage: [![Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1ZvU9ZSHJKSeJyH5bgxo_A-GSVIUcSt2E?usp=sharing) Notebook for Imagenette evaluation: [![Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1-Duq0YNtjzOnmuXCYo-5DDiOzeCItXpN?usp=sharing) ## The team This model and the example pipeline were created by Eugenia Iofinova, Michael Goin, Chris Wendler, and Dan Alistarh. Special thanks to Abhinav Agarwalla and Alexandre Marques for technical support with parts of the project. ## Setup for usage First, install DeepSparse with extensions for CLIP: ``` pip install deepsparse-nightly[clip]>=1.7.0.20231210 ``` Download some test images of a church, a dog, and elephants: ``` wget -O basilica.jpg https://raw.githubusercontent.com/neuralmagic/deepsparse/main/src/deepsparse/yolo/sample_images/basilica.jpg wget -O buddy.jpeg https://raw.githubusercontent.com/neuralmagic/deepsparse/main/tests/deepsparse/pipelines/sample_images/buddy.jpeg wget -O thailand.jpg https://raw.githubusercontent.com/neuralmagic/deepsparse/main/src/deepsparse/yolact/sample_images/thailand.jpg ``` For this model there is a second input that is the length of tokens, so run this input override code before making a text pipeline: ```python import numpy as np from deepsparse.clip import CLIPTextPipeline def custom_process_inputs(self, inputs): if not isinstance(inputs.text, list): inputs.text = [inputs.text] if not isinstance(inputs.text[0], str): return inputs.text tokens = [np.array(t).astype(np.int32) for t in self.tokenizer(inputs.text)] tokens = np.stack(tokens, axis=0) tokens_lengths = np.array(tokens.shape[0] * [tokens.shape[1] - 1]) return [tokens, tokens_lengths] # This overrides the process_inputs function globally for all CLIPTextPipeline classes CLIPTextPipeline.process_inputs = custom_process_inputs ``` ## Text embedding pipeline Here is an example of how to create and use a [DeepSparse pipeline for text embeddings](https://github.com/neuralmagic/deepsparse/blob/main/src/deepsparse/clip/text_pipeline.py). ```python from deepsparse import Pipeline from huggingface_hub import snapshot_download # Download the model from HF model_folder = snapshot_download(repo_id="neuralmagic/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K-quant-ds") text_embed_pipeline = Pipeline.create(task="clip_text", model_path=model_folder + "/textual.onnx") text = ["ice cream", "an elephant", "a dog", "a building", "a church"] embeddings = text_embed_pipeline(text=text).text_embeddings for i in range(len(embeddings)): print(embeddings[i].shape) print(embeddings[i]) ``` ## Image embedding pipeline Here is an example of how to create and use a [DeepSparse pipeline for image embeddings](https://github.com/neuralmagic/deepsparse/blob/main/src/deepsparse/clip/visual_pipeline.py). ```python from deepsparse import Pipeline from huggingface_hub import snapshot_download # Download the model from HF model_folder = snapshot_download(repo_id="neuralmagic/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K-quant-ds") image_embed_pipeline = Pipeline.create(task="clip_visual", model_path=model_folder + "/visual.onnx") images = ["basilica.jpg", "buddy.jpeg", "thailand.jpg"] embeddings = image_embed_pipeline(images=images).image_embeddings for i in range(len(embeddings)): print(embeddings[i].shape) print(embeddings[i]) ``` ## Zero-shot image classification pipeline Since CLIP trained both the text and image embedding models in tandem, we can generate embeddings for both and relate them together without retraining. Here is an example of how to create and use a [DeepSparse pipeline for zero-shot image classification](https://github.com/neuralmagic/deepsparse/blob/main/src/deepsparse/clip/zeroshot_pipeline.py). ```python from deepsparse import Pipeline from deepsparse.clip import ( CLIPTextInput, CLIPVisualInput, CLIPZeroShotInput ) from huggingface_hub import snapshot_download # Download the model from HF model_folder = snapshot_download(repo_id="neuralmagic/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K-quant-ds") possible_classes = ["ice cream", "an elephant", "a dog", "a building", "a church"] images = ["basilica.jpg", "buddy.jpeg", "thailand.jpg"] # Load the model into DeepSparse pipeline = Pipeline.create( task="clip_zeroshot", visual_model_path=model_folder + "/visual.onnx", text_model_path=model_folder + "/textual.onnx" ) # Infer output = pipeline( image=CLIPVisualInput(images=images), text=CLIPTextInput(text=possible_classes), ).text_scores for i in range(len(output)): prediction = possible_classes[np.argmax(output[i])] print(f"Image {images[i]} is a picture of {prediction}") """ Image basilica.jpg is a picture of a church Image buddy.jpeg is a picture of a dog Image thailand.jpg is a picture of an elephant """ ```
Ricardo54321/q-FrozenLake-v1-4x4-noSlippery	Ricardo54321	2024-01-12T22:35:28Z	0	0	null	[ "FrozenLake-v1-4x4-no_slippery", "q-learning", "reinforcement-learning", "custom-implementation", "model-index", "region:us" ]	reinforcement-learning	2024-01-12T22:35:17Z	--- tags: - FrozenLake-v1-4x4-no_slippery - q-learning - reinforcement-learning - custom-implementation model-index: - name: q-FrozenLake-v1-4x4-noSlippery results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: FrozenLake-v1-4x4-no_slippery type: FrozenLake-v1-4x4-no_slippery metrics: - type: mean_reward value: 1.00 +/- 0.00 name: mean_reward verified: false --- # Q-Learning Agent playing1 FrozenLake-v1 This is a trained model of a Q-Learning agent playing FrozenLake-v1 . ## Usage ```python model = load_from_hub(repo_id="Ricardo54321/q-FrozenLake-v1-4x4-noSlippery", filename="q-learning.pkl") # Don't forget to check if you need to add additional attributes (is_slippery=False etc) env = gym.make(model["env_id"]) ```
jysssacc/opt-350m_IA3_lr0.05_bs10_epoch5_wd0.01	jysssacc	2024-01-12T22:22:30Z	0	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:facebook/opt-350m", "base_model:adapter:facebook/opt-350m", "license:other", "region:us" ]	null	2024-01-12T22:22:04Z	--- license: other library_name: peft tags: - generated_from_trainer base_model: facebook/opt-350m model-index: - name: opt-350m_IA3_lr0.05_bs10_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # opt-350m_IA3_lr0.05_bs10_epoch5_wd0.01 This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.6540 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.05 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| No log \| 1.0 \| 63 \| 3.4823 \| \| 3.6838 \| 2.0 \| 126 \| 3.4466 \| \| 3.6838 \| 3.0 \| 189 \| 3.4563 \| \| 3.1836 \| 4.0 \| 252 \| 3.5795 \| \| 2.8258 \| 5.0 \| 315 \| 3.6540 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
MariaK/peft-starcoder-lora	MariaK	2024-01-12T22:15:34Z	0	0	null	[ "tensorboard", "safetensors", "generated_from_trainer", "base_model:bigcode/starcoderbase-1b", "base_model:finetune:bigcode/starcoderbase-1b", "license:bigcode-openrail-m", "region:us" ]	null	2023-12-14T16:01:10Z	--- license: bigcode-openrail-m base_model: bigcode/starcoderbase-1b tags: - generated_from_trainer model-index: - name: peft-starcoder-lora results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # peft-starcoder-lora This model is a fine-tuned version of [bigcode/starcoderbase-1b](https://huggingface.co/bigcode/starcoderbase-1b) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 5.9439 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0005 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine - lr_scheduler_warmup_steps: 30 - training_steps: 2000 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| 5.8019 \| 0.05 \| 100 \| 8.1006 \| \| 5.9737 \| 0.1 \| 200 \| 7.2985 \| \| 5.1109 \| 0.15 \| 300 \| 6.3542 \| \| 5.0956 \| 0.2 \| 400 \| 6.2918 \| \| 5.2913 \| 0.25 \| 500 \| 6.1454 \| \| 4.6567 \| 0.3 \| 600 \| 6.0157 \| \| 5.1439 \| 0.35 \| 700 \| 6.0052 \| \| 5.2159 \| 0.4 \| 800 \| 6.0141 \| \| 4.7227 \| 0.45 \| 900 \| 5.9708 \| \| 5.2372 \| 0.5 \| 1000 \| 5.9662 \| \| 5.5667 \| 0.55 \| 1100 \| 6.1102 \| \| 4.6813 \| 0.6 \| 1200 \| 5.9753 \| \| 5.1002 \| 0.65 \| 1300 \| 5.9829 \| \| 5.327 \| 0.7 \| 1400 \| 5.9327 \| \| 4.6781 \| 0.75 \| 1500 \| 5.9415 \| \| 5.2732 \| 0.8 \| 1600 \| 5.9264 \| \| 5.2995 \| 0.85 \| 1700 \| 5.9178 \| \| 4.5842 \| 0.9 \| 1800 \| 5.9290 \| \| 5.2401 \| 0.95 \| 1900 \| 5.9506 \| \| 5.1396 \| 1.0 \| 2000 \| 5.9439 \| ### Framework versions - Transformers 4.35.2 - Pytorch 2.1.0+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
dan0102dan/ohuenko	dan0102dan	2024-01-12T22:15:31Z	1	0	diffusers	[ "diffusers", "text-to-image", "autotrain", "base_model:stabilityai/stable-diffusion-xl-base-1.0", "base_model:finetune:stabilityai/stable-diffusion-xl-base-1.0", "region:us" ]	text-to-image	2024-01-12T12:28:56Z	--- base_model: stabilityai/stable-diffusion-xl-base-1.0 instance_prompt: arts sks tags: - text-to-image - diffusers - autotrain inference: true --- # DreamBooth trained by AutoTrain Text encoder was not trained.
TheBloke/PiVoT-SUS-RP-AWQ	TheBloke	2024-01-12T22:13:01Z	7	1	transformers	[ "transformers", "safetensors", "llama", "text-generation", "base_model:maywell/PiVoT-SUS-RP", "base_model:quantized:maywell/PiVoT-SUS-RP", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "4-bit", "awq", "region:us" ]	text-generation	2024-01-12T18:57:15Z	--- base_model: maywell/PiVoT-SUS-RP inference: false license: apache-2.0 model_creator: Jeonghwan Park model_name: Pivot SUS RP model_type: yi prompt_template: '{prompt} ' quantized_by: TheBloke --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Pivot SUS RP - AWQ - Model creator: [Jeonghwan Park](https://huggingface.co/maywell) - Original model: [Pivot SUS RP](https://huggingface.co/maywell/PiVoT-SUS-RP) <!-- description start --> ## Description This repo contains AWQ model files for [Jeonghwan Park's Pivot SUS RP](https://huggingface.co/maywell/PiVoT-SUS-RP). ### About AWQ AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference with equivalent or better quality compared to the most commonly used GPTQ settings. AWQ models are currently supported on Linux and Windows, with NVidia GPUs only. macOS users: please use GGUF models instead. It is supported by: - [Text Generation Webui](https://github.com/oobabooga/text-generation-webui) - using Loader: AutoAWQ - [vLLM](https://github.com/vllm-project/vllm) - version 0.2.2 or later for support for all model types. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later, from any code or client that supports Transformers - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) - for use from Python code <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/PiVoT-SUS-RP-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/PiVoT-SUS-RP-GGUF) * [Jeonghwan Park's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/maywell/PiVoT-SUS-RP) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: Unknown ``` {prompt} ``` <!-- prompt-template end --> <!-- README_AWQ.md-provided-files start --> ## Provided files, and AWQ parameters I currently release 128g GEMM models only. The addition of group_size 32 models, and GEMV kernel models, is being actively considered. Models are released as sharded safetensors files. \| Branch \| Bits \| GS \| AWQ Dataset \| Seq Len \| Size \| \| ------ \| ---- \| -- \| ----------- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/PiVoT-SUS-RP-AWQ/tree/main) \| 4 \| 128 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 8192 \| 19.23 GB <!-- README_AWQ.md-provided-files end --> <!-- README_AWQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/PiVoT-SUS-RP-AWQ`. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `PiVoT-SUS-RP-AWQ` 7. Select Loader: AutoAWQ. 8. Click Load, and the model will load and is now ready for use. 9. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. 10. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_AWQ.md-text-generation-webui end --> <!-- README_AWQ.md-use-from-vllm start --> ## Multi-user inference server: vLLM Documentation on installing and using vLLM [can be found here](https://vllm.readthedocs.io/en/latest/). - Please ensure you are using vLLM version 0.2 or later. - When using vLLM as a server, pass the `--quantization awq` parameter. For example: ```shell python3 -m vllm.entrypoints.api_server --model TheBloke/PiVoT-SUS-RP-AWQ --quantization awq --dtype auto ``` - When using vLLM from Python code, again set `quantization=awq`. For example: ```python from vllm import LLM, SamplingParams prompts = [ "Tell me about AI", "Write a story about llamas", "What is 291 - 150?", "How much wood would a woodchuck chuck if a woodchuck could chuck wood?", ] prompt_template=f'''{prompt} ''' prompts = [prompt_template.format(prompt=prompt) for prompt in prompts] sampling_params = SamplingParams(temperature=0.8, top_p=0.95) llm = LLM(model="TheBloke/PiVoT-SUS-RP-AWQ", quantization="awq", dtype="auto") outputs = llm.generate(prompts, sampling_params) # Print the outputs. for output in outputs: prompt = output.prompt generated_text = output.outputs[0].text print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}") ``` <!-- README_AWQ.md-use-from-vllm start --> <!-- README_AWQ.md-use-from-tgi start --> ## Multi-user inference server: Hugging Face Text Generation Inference (TGI) Use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/PiVoT-SUS-RP-AWQ --port 3000 --quantize awq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires [huggingface-hub](https://github.com/huggingface/huggingface_hub) 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''{prompt} ''' client = InferenceClient(endpoint_url) response = client.text_generation(prompt, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1) print(f"Model output: ", response) ``` <!-- README_AWQ.md-use-from-tgi end --> <!-- README_AWQ.md-use-from-python start --> ## Inference from Python code using Transformers ### Install the necessary packages - Requires: [Transformers](https://huggingface.co/docs/transformers) 4.35.0 or later. - Requires: [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) 0.1.6 or later. ```shell pip3 install --upgrade "autoawq>=0.1.6" "transformers>=4.35.0" ``` Note that if you are using PyTorch 2.0.1, the above AutoAWQ command will automatically upgrade you to PyTorch 2.1.0. If you are using CUDA 11.8 and wish to continue using PyTorch 2.0.1, instead run this command: ```shell pip3 install https://github.com/casper-hansen/AutoAWQ/releases/download/v0.1.6/autoawq-0.1.6+cu118-cp310-cp310-linux_x86_64.whl ``` If you have problems installing [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) using the pre-built wheels, install it from source instead: ```shell pip3 uninstall -y autoawq git clone https://github.com/casper-hansen/AutoAWQ cd AutoAWQ pip3 install . ``` ### Transformers example code (requires Transformers 4.35.0 and later) ```python from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer model_name_or_path = "TheBloke/PiVoT-SUS-RP-AWQ" tokenizer = AutoTokenizer.from_pretrained(model_name_or_path) model = AutoModelForCausalLM.from_pretrained( model_name_or_path, low_cpu_mem_usage=True, device_map="cuda:0" ) # Using the text streamer to stream output one token at a time streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True) prompt = "Tell me about AI" prompt_template=f'''{prompt} ''' # Convert prompt to tokens tokens = tokenizer( prompt_template, return_tensors='pt' ).input_ids.cuda() generation_params = { "do_sample": True, "temperature": 0.7, "top_p": 0.95, "top_k": 40, "max_new_tokens": 512, "repetition_penalty": 1.1 } # Generate streamed output, visible one token at a time generation_output = model.generate( tokens, streamer=streamer, generation_params ) # Generation without a streamer, which will include the prompt in the output generation_output = model.generate( tokens, generation_params ) # Get the tokens from the output, decode them, print them token_output = generation_output[0] text_output = tokenizer.decode(token_output) print("model.generate output: ", text_output) # Inference is also possible via Transformers' pipeline from transformers import pipeline pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, *generation_params ) pipe_output = pipe(prompt_template)[0]['generated_text'] print("pipeline output: ", pipe_output) ``` <!-- README_AWQ.md-use-from-python end --> <!-- README_AWQ.md-compatibility start --> ## Compatibility The files provided are tested to work with: - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) using `Loader: AutoAWQ`. - [vLLM](https://github.com/vllm-project/vllm) version 0.2.0 and later. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) version 1.1.0 and later. - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later. - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) version 0.1.1 and later. <!-- README_AWQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Jeonghwan Park's Pivot SUS RP
togethercomputer/m2-bert-80M-32k-retrieval	togethercomputer	2024-01-12T21:49:00Z	439	125	transformers	[ "transformers", "pytorch", "m2_bert", "text-classification", "sentence-similarity", "custom_code", "en", "arxiv:2310.12109", "license:apache-2.0", "autotrain_compatible", "region:us" ]	sentence-similarity	2023-11-04T03:09:52Z	--- license: apache-2.0 language: - en pipeline_tag: sentence-similarity inference: false --- # Monarch Mixer-BERT An 80M checkpoint of M2-BERT, pretrained with sequence length 32768, and it has been fine-tuned for long-context retrieval. Check out the paper [Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture](https://arxiv.org/abs/2310.12109) and our [blog post]() on retrieval for more on how we trained this model for long sequence. This model was trained by Jon Saad-Falcon, Dan Fu, and Simran Arora. Check out our [GitHub](https://github.com/HazyResearch/m2/tree/main) for instructions on how to download and fine-tune it! ## How to use You can load this model using Hugging Face `AutoModel`: ```python from transformers import AutoModelForSequenceClassification model = AutoModelForSequenceClassification.from_pretrained( "togethercomputer/m2-bert-80M-32k-retrieval", trust_remote_code=True ) ``` You should expect to see a large error message about unused parameters for FlashFFTConv. If you'd like to load the model with FlashFFTConv, you can check out our [GitHub](https://github.com/HazyResearch/m2/tree/main). This model generates embeddings for retrieval. The embeddings have a dimensionality of 768: ```python from transformers import AutoTokenizer, AutoModelForSequenceClassification max_seq_length = 32768 testing_string = "Every morning, I make a cup of coffee to start my day." model = AutoModelForSequenceClassification.from_pretrained( "togethercomputer/m2-bert-80M-32k-retrieval", trust_remote_code=True ) tokenizer = AutoTokenizer.from_pretrained( "bert-base-uncased", model_max_length=max_seq_length ) input_ids = tokenizer( [testing_string], return_tensors="pt", padding="max_length", return_token_type_ids=False, truncation=True, max_length=max_seq_length ) outputs = model(**input_ids) embeddings = outputs['sentence_embedding'] ``` You can also get embeddings from this model using the Together API as follows (you can find your API key [here](https://api.together.xyz/settings/api-keys)): ```python import os import requests def generate_together_embeddings(text: str, model_api_string: str, api_key: str): url = "https://api.together.xyz/api/v1/embeddings" headers = { "accept": "application/json", "content-type": "application/json", "Authorization": f"Bearer {api_key}" } session = requests.Session() response = session.post( url, headers=headers, json={ "input": text, "model": model_api_string } ) if response.status_code != 200: raise ValueError(f"Request failed with status code {response.status_code}: {response.text}") return response.json()['data'][0]['embedding'] print(generate_together_embeddings( 'Hello world', 'togethercomputer/m2-bert-80M-32k-retrieval', os.environ['TOGETHER_API_KEY'])[:10] ) ``` ## Acknowledgments Alycia Lee helped with AutoModel support. ## Citation If you use this model, or otherwise found our work valuable, you can cite us as follows: ``` @inproceedings{fu2023monarch, title={Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture}, author={Fu, Daniel Y and Arora, Simran and Grogan, Jessica and Johnson, Isys and Eyuboglu, Sabri and Thomas, Armin W and Spector, Benjamin and Poli, Michael and Rudra, Atri and R{\'e}, Christopher}, booktitle={Advances in Neural Information Processing Systems}, year={2023} } ```
alexsherstinsky/zephyr-7b-beta-based-finetuned-using-ludwig-with-jigsaw-T4-4bit-notmerged	alexsherstinsky	2024-01-12T21:43:42Z	2	0	peft	[ "peft", "safetensors", "arxiv:1910.09700", "base_model:HuggingFaceH4/zephyr-7b-beta", "base_model:adapter:HuggingFaceH4/zephyr-7b-beta", "region:us" ]	null	2024-01-10T21:35:02Z	--- library_name: peft base_model: HuggingFaceH4/zephyr-7b-beta --- # Model Card for Model ID <!-- Provide a quick summary of what the model is/does. --> ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> - Developed by: [More Information Needed] - Funded by [optional]: [More Information Needed] - Shared by [optional]: [More Information Needed] - Model type: [More Information Needed] - Language(s) (NLP): [More Information Needed] - License: [More Information Needed] - Finetuned from model [optional]: [More Information Needed] ### Model Sources [optional] <!-- Provide the basic links for the model. --> - Repository: [More Information Needed] - Paper [optional]: [More Information Needed] - Demo [optional]: [More Information Needed] ## Uses <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> [More Information Needed] ### Downstream Use [optional] <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app --> [More Information Needed] ### Out-of-Scope Use <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> [More Information Needed] ## Bias, Risks, and Limitations <!-- This section is meant to convey both technical and sociotechnical limitations. --> [More Information Needed] ### Recommendations <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. --> Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. ## How to Get Started with the Model Use the code below to get started with the model. [More Information Needed] ## Training Details ### Training Data <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> [More Information Needed] ### Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> #### Preprocessing [optional] [More Information Needed] #### Training Hyperparameters - Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision --> #### Speeds, Sizes, Times [optional] <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. --> [More Information Needed] ## Evaluation <!-- This section describes the evaluation protocols and provides the results. --> ### Testing Data, Factors & Metrics #### Testing Data <!-- This should link to a Dataset Card if possible. --> [More Information Needed] #### Factors <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. --> [More Information Needed] #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> [More Information Needed] ### Results [More Information Needed] #### Summary ## Model Examination [optional] <!-- Relevant interpretability work for the model goes here --> [More Information Needed] ## Environmental Impact <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly --> Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - Hardware Type: [More Information Needed] - Hours used: [More Information Needed] - Cloud Provider: [More Information Needed] - Compute Region: [More Information Needed] - Carbon Emitted: [More Information Needed] ## Technical Specifications [optional] ### Model Architecture and Objective [More Information Needed] ### Compute Infrastructure [More Information Needed] #### Hardware [More Information Needed] #### Software [More Information Needed] ## Citation [optional] <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> BibTeX: [More Information Needed] APA: [More Information Needed] ## Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> [More Information Needed] ## More Information [optional] [More Information Needed] ## Model Card Authors [optional] [More Information Needed] ## Model Card Contact [More Information Needed] ### Framework versions - PEFT 0.7.1
xaviviro/ElChato-0.1-1.1b-GGUF	xaviviro	2024-01-12T21:38:18Z	65	2	null	[ "gguf", "es", "license:apache-2.0", "endpoints_compatible", "region:us", "conversational" ]	null	2024-01-04T07:08:20Z	--- license: apache-2.0 language: - es model_creator: xaviviro model_name: PetitXat-0.1-1.1b prompt_template: <\|system\|>\n{system}</s>\n<\|user\|>{instruction}</s>\n<\|assistant\|>\n --- # ElChato 1.1B: El modelo más pequeño de chat en lengua castellana ![ElChato](elchato.png) ElChato és el modelo más pequeño de chat en lengua castellana. És el resultado de finetunear el modelo [TinyLlama-1.1B-Chat-v1.0](/TinyLlama/TinyLlama-1.1B-Chat-v1.0) con las instrucciones de [OpenAssistant v2](/datasets/OpenAssistant/oasst2) traducidas automáticamente al castellano con recursos de [Helsinki-NLP](/Helsinki-NLP) y tratadas en formato ChatGLM3. ## Formato del prompt ``` <\|system\|> Eres un buen asistente</s> <\|user\|> ¿Quién fue Isaac Newton?</s> <\|assistant\|> ```
kim-sha/sql-sodabot-v1.0	kim-sha	2024-01-12T21:36:02Z	92	0	transformers	[ "transformers", "safetensors", "t5", "text2text-generation", "generated_from_trainer", "base_model:Salesforce/codet5-small", "base_model:finetune:Salesforce/codet5-small", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text2text-generation	2024-01-12T18:05:46Z	--- license: apache-2.0 base_model: Salesforce/codet5-small tags: - generated_from_trainer model-index: - name: sodabot-sql-sm results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # sql-sodabot-v1.0 This encoder-decoder model is a descendent of [Salesforce/codet5-small](https://huggingface.co/Salesforce/codet5-small), fine-tuned on a modified version of [b-mc2/sql-create-context](https://huggingface.co/datasets/b-mc2/sql-create-context) data. The original [CodeT5](https://github.com/salesforce/CodeT5) was published by Salesfoce Research as an "AI-powered coding assistant to boost the productivity of software developers". The goal of this project is to apply transfer learning in order to appropriate this model for text-to-SQL applications, specifically in the context of generating Socrata SQL ([SoQL](https://dev.socrata.com/docs/queries/)) queries that can be executed on the Socrata Open Data API (e.g., to analyze [NYC Open Data](https://opendata.cityofnewyork.us)). ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 64 - eval_batch_size: 64 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Framework versions - Transformers 4.35.2 - Pytorch 2.1.0+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
Denox05/cc	Denox05	2024-01-12T21:23:15Z	0	0	null	[ "license:other", "region:us" ]	null	2024-01-12T21:22:05Z	--- license: other license_name: rv license_link: LICENSE ---
Polu/ppo-SnowballTarget	Polu	2024-01-12T21:15:16Z	0	0	ml-agents	[ "ml-agents", "tensorboard", "onnx", "SnowballTarget", "deep-reinforcement-learning", "reinforcement-learning", "ML-Agents-SnowballTarget", "region:us" ]	reinforcement-learning	2024-01-12T21:15:08Z	--- library_name: ml-agents tags: - SnowballTarget - deep-reinforcement-learning - reinforcement-learning - ML-Agents-SnowballTarget --- # ppo Agent playing SnowballTarget This is a trained model of a ppo agent playing SnowballTarget using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents). ## Usage (with ML-Agents) The Documentation: https://unity-technologies.github.io/ml-agents/ML-Agents-Toolkit-Documentation/ We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub: - A short tutorial where you teach Huggy the Dog 🐶 to fetch the stick and then play with him directly in your browser: https://huggingface.co/learn/deep-rl-course/unitbonus1/introduction - A longer tutorial to understand how works ML-Agents: https://huggingface.co/learn/deep-rl-course/unit5/introduction ### Resume the training ```bash mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume ``` ### Watch your Agent play You can watch your agent playing directly in your browser 1. If the environment is part of ML-Agents official environments, go to https://huggingface.co/unity 2. Step 1: Find your model_id: Polu/ppo-SnowballTarget 3. Step 2: Select your .nn /.onnx file 4. Click on Watch the agent play 👀
yaizaa/winxclub	yaizaa	2024-01-12T20:59:19Z	0	1	fastai	[ "fastai", "region:us" ]	null	2024-01-12T20:10:31Z	--- tags: - fastai --- # Amazing! 🥳 Congratulations on hosting your fastai model on the Hugging Face Hub! # Some next steps 1. Fill out this model card with more information (see the template below and the [documentation here](https://huggingface.co/docs/hub/model-repos))! 2. Create a demo in Gradio or Streamlit using 🤗 Spaces ([documentation here](https://huggingface.co/docs/hub/spaces)). 3. Join the fastai community on the [Fastai Discord](https://discord.com/invite/YKrxeNn)! Greetings fellow fastlearner 🤝! Don't forget to delete this content from your model card. --- # Model card ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed
TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ	TheBloke	2024-01-12T20:55:59Z	12	2	transformers	[ "transformers", "safetensors", "mixtral", "text-generation", "conversational", "base_model:NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss", "base_model:quantized:NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss", "license:cc-by-nc-4.0", "autotrain_compatible", "text-generation-inference", "4-bit", "awq", "region:us" ]	text-generation	2024-01-12T20:14:59Z	--- base_model: NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss inference: false license: cc-by-nc-4.0 model_creator: NeverSleep model_name: Noromaid V0.4 Mixtral Instruct 8X7B ZLoss model_type: mixtral prompt_template: '<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ' quantized_by: TheBloke --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Noromaid V0.4 Mixtral Instruct 8X7B ZLoss - AWQ - Model creator: [NeverSleep](https://huggingface.co/NeverSleep) - Original model: [Noromaid V0.4 Mixtral Instruct 8X7B ZLoss](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss) <!-- description start --> ## Description This repo contains AWQ model files for [NeverSleep's Noromaid V0.4 Mixtral Instruct 8X7B ZLoss](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss). These files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/). MIXTRAL AWQ This is a Mixtral AWQ model. For AutoAWQ inference, please install AutoAWQ 0.1.8 or later. Support via Transformers is also available, but currently requires installing Transformers from Github: `pip3 install git+https://github.com/huggingface/transformers.git` vLLM: version 0.2.6 is confirmed to support Mixtral AWQs. TGI: I tested version 1.3.3 and it loaded the model fine, but I was not able to get any output back. Further testing/debug is required. (Let me know if you get it working!) ### About AWQ AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference with equivalent or better quality compared to the most commonly used GPTQ settings. AWQ models are currently supported on Linux and Windows, with NVidia GPUs only. macOS users: please use GGUF models instead. AWQ models are supported by (note that not all of these may support Mixtral models yet - see above): - [Text Generation Webui](https://github.com/oobabooga/text-generation-webui) - using Loader: AutoAWQ - [vLLM](https://github.com/vllm-project/vllm) - version 0.2.2 or later for support for all model types. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later, from any code or client that supports Transformers - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) - for use from Python code <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF) * [NeverSleep's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: ChatML ``` <\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ``` <!-- prompt-template end --> <!-- README_AWQ.md-provided-files start --> ## Provided files, and AWQ parameters I currently release 128g GEMM models only. The addition of group_size 32 models, and GEMV kernel models, is being actively considered. Models are released as sharded safetensors files. \| Branch \| Bits \| GS \| AWQ Dataset \| Seq Len \| Size \| \| ------ \| ---- \| -- \| ----------- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ/tree/main) \| 4 \| 128 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 8192 \| 24.65 GB <!-- README_AWQ.md-provided-files end --> <!-- README_AWQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ`. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ` 7. Select Loader: AutoAWQ. 8. Click Load, and the model will load and is now ready for use. 9. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. 10. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_AWQ.md-text-generation-webui end --> <!-- README_AWQ.md-use-from-vllm start --> ## Multi-user inference server: vLLM Documentation on installing and using vLLM [can be found here](https://vllm.readthedocs.io/en/latest/). - Please ensure you are using vLLM version 0.2 or later. - When using vLLM as a server, pass the `--quantization awq` parameter. For example: ```shell python3 -m vllm.entrypoints.api_server --model TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ --quantization awq --dtype auto ``` - When using vLLM from Python code, again set `quantization=awq`. For example: ```python from vllm import LLM, SamplingParams prompts = [ "Tell me about AI", "Write a story about llamas", "What is 291 - 150?", "How much wood would a woodchuck chuck if a woodchuck could chuck wood?", ] prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' prompts = [prompt_template.format(prompt=prompt) for prompt in prompts] sampling_params = SamplingParams(temperature=0.8, top_p=0.95) llm = LLM(model="TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ", quantization="awq", dtype="auto") outputs = llm.generate(prompts, sampling_params) # Print the outputs. for output in outputs: prompt = output.prompt generated_text = output.outputs[0].text print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}") ``` <!-- README_AWQ.md-use-from-vllm start --> <!-- README_AWQ.md-use-from-tgi start --> ## Multi-user inference server: Hugging Face Text Generation Inference (TGI) Use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ --port 3000 --quantize awq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires [huggingface-hub](https://github.com/huggingface/huggingface_hub) 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' client = InferenceClient(endpoint_url) response = client.text_generation(prompt, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1) print(f"Model output: ", response) ``` <!-- README_AWQ.md-use-from-tgi end --> <!-- README_AWQ.md-use-from-python start --> ## Inference from Python code using Transformers ### Install the necessary packages - Requires: [Transformers](https://huggingface.co/docs/transformers) 4.35.0 or later. - Requires: [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) 0.1.6 or later. ```shell pip3 install --upgrade "autoawq>=0.1.6" "transformers>=4.35.0" ``` Note that if you are using PyTorch 2.0.1, the above AutoAWQ command will automatically upgrade you to PyTorch 2.1.0. If you are using CUDA 11.8 and wish to continue using PyTorch 2.0.1, instead run this command: ```shell pip3 install https://github.com/casper-hansen/AutoAWQ/releases/download/v0.1.6/autoawq-0.1.6+cu118-cp310-cp310-linux_x86_64.whl ``` If you have problems installing [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) using the pre-built wheels, install it from source instead: ```shell pip3 uninstall -y autoawq git clone https://github.com/casper-hansen/AutoAWQ cd AutoAWQ pip3 install . ``` ### Transformers example code (requires Transformers 4.35.0 and later) ```python from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer model_name_or_path = "TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ" tokenizer = AutoTokenizer.from_pretrained(model_name_or_path) model = AutoModelForCausalLM.from_pretrained( model_name_or_path, low_cpu_mem_usage=True, device_map="cuda:0" ) # Using the text streamer to stream output one token at a time streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True) prompt = "Tell me about AI" prompt_template=f'''<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ''' # Convert prompt to tokens tokens = tokenizer( prompt_template, return_tensors='pt' ).input_ids.cuda() generation_params = { "do_sample": True, "temperature": 0.7, "top_p": 0.95, "top_k": 40, "max_new_tokens": 512, "repetition_penalty": 1.1 } # Generate streamed output, visible one token at a time generation_output = model.generate( tokens, streamer=streamer, generation_params ) # Generation without a streamer, which will include the prompt in the output generation_output = model.generate( tokens, generation_params ) # Get the tokens from the output, decode them, print them token_output = generation_output[0] text_output = tokenizer.decode(token_output) print("model.generate output: ", text_output) # Inference is also possible via Transformers' pipeline from transformers import pipeline pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, *generation_params ) pipe_output = pipe(prompt_template)[0]['generated_text'] print("pipeline output: ", pipe_output) ``` <!-- README_AWQ.md-use-from-python end --> <!-- README_AWQ.md-compatibility start --> ## Compatibility The files provided are tested to work with: - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) using `Loader: AutoAWQ`. - [vLLM](https://github.com/vllm-project/vllm) version 0.2.0 and later. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) version 1.1.0 and later. - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later. - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) version 0.1.1 and later. <!-- README_AWQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: NeverSleep's Noromaid V0.4 Mixtral Instruct 8X7B ZLoss ![image/png](https://cdn-uploads.huggingface.co/production/uploads/630dfb008df86f1e5becadc3/vwcJfOnL-2QDJ0ShfxRJ5.png) --- # Disclaimer: ## This model is experimental, do not expect everything to work. This model uses the Chatml prompting format --- Beeg noromaid on *steroids*. Suitable for RP, ERP. This model was trained on the Zloss fork of Charles, and should fix issue the model had. Use Chatml prompt format, but not the special token. The reason is that Axolotl merge the finetune with the base model at 1.0 weight basically, but this is too much, so I use another script available [HERE](https://github.com/DocShotgun/LLM-notebooks/blob/main/weighted-lora-merge.ipynb) to merge with less weight, sadly, it don't take the special Chatml token. It's like Orca2 for the matter. ## Credits: - Undi - IkariDev <!-- description start --> ## Description <!-- [Recommended settings - contributed by localfultonextractor](https://files.catbox.moe/ue0tja.json) --> This repo contains FP16 files of Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss. [FP16 - by IkariDev and Undi](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss) <!-- [GGUF - By TheBloke](https://huggingface.co/TheBloke/Athena-v4-GGUF)--> <!-- [GPTQ - By TheBloke](https://huggingface.co/TheBloke/Athena-v4-GPTQ)--> <!-- [exl2[8bpw-8h] - by AzureBlack](https://huggingface.co/AzureBlack/Echidna-13b-v0.3-8bpw-8h-exl2)--> <!-- [AWQ - By TheBloke](https://huggingface.co/TheBloke/Athena-v4-AWQ)--> <!-- [fp16 - by IkariDev+Undi95](https://huggingface.co/IkariDev/Athena-v4)--> [GGUF - by IkariDev and Undi](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF) <!-- [OLD(GGUF - by IkariDev+Undi95)](https://huggingface.co/IkariDev/Athena-v4-GGUF)--> ## Ratings: Note: We have permission of all users to upload their ratings, we DONT screenshot random reviews without asking if we can put them here! No ratings yet! If you want your rating to be here, send us a message over on DC and we'll put up a screenshot of it here. DC name is "ikaridev" and "undi". <!-- description end --> <!-- prompt-template start --> ### Prompt format: Chatml ``` <\|im_start\|>system {sysprompt}<\|im_end\|> <\|im_start\|>user {input}<\|im_end\|> <\|im_start\|>assistant {output}<\|im_end\|> ``` ## Datasets used: - Aesir 1, 2 & 3 modified by us, credit to ([MinervaAI](https://huggingface.co/MinervaAI) / [Gryphe](https://huggingface.co/Gryphe)) - [LimaRP-20231109](https://huggingface.co/datasets/lemonilia/LimaRP) ([Lemonilia](https://huggingface.co/lemonilia)) - [ToxicQAFinal](https://huggingface.co/datasets/NobodyExistsOnTheInternet/ToxicQAFinal) ([NobodyExistsOnTheInternet](https://huggingface.co/NobodyExistsOnTheInternet) - [No-robots-ShareGPT](https://huggingface.co/datasets/Doctor-Shotgun/no-robots-sharegpt) ([Doctor-Shotgun](https://huggingface.co/Doctor-Shotgun)) ## Others Undi: If you want to support me, you can [here](https://ko-fi.com/undiai). IkariDev: Visit my [retro/neocities style website](https://ikaridevgit.github.io/) please kek
Ricardo54321/Huggy	Ricardo54321	2024-01-12T20:54:10Z	20	0	ml-agents	[ "ml-agents", "tensorboard", "onnx", "Huggy", "deep-reinforcement-learning", "reinforcement-learning", "ML-Agents-Huggy", "region:us" ]	reinforcement-learning	2024-01-12T20:53:54Z	--- library_name: ml-agents tags: - Huggy - deep-reinforcement-learning - reinforcement-learning - ML-Agents-Huggy --- # ppo Agent playing Huggy This is a trained model of a ppo agent playing Huggy using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents). ## Usage (with ML-Agents) The Documentation: https://unity-technologies.github.io/ml-agents/ML-Agents-Toolkit-Documentation/ We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub: - A short tutorial where you teach Huggy the Dog 🐶 to fetch the stick and then play with him directly in your browser: https://huggingface.co/learn/deep-rl-course/unitbonus1/introduction - A longer tutorial to understand how works ML-Agents: https://huggingface.co/learn/deep-rl-course/unit5/introduction ### Resume the training ```bash mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume ``` ### Watch your Agent play You can watch your agent playing directly in your browser 1. If the environment is part of ML-Agents official environments, go to https://huggingface.co/unity 2. Step 1: Find your model_id: Ricardo54321/Huggy 3. Step 2: Select your .nn /.onnx file 4. Click on Watch the agent play 👀
jysssacc/opt-350m_lora_lr0.005_bs10_epoch5_wd0.01	jysssacc	2024-01-12T20:47:44Z	1	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:facebook/opt-350m", "base_model:adapter:facebook/opt-350m", "license:other", "region:us" ]	null	2024-01-12T20:47:06Z	--- license: other library_name: peft tags: - generated_from_trainer base_model: facebook/opt-350m model-index: - name: opt-350m_lora_lr0.005_bs10_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # opt-350m_lora_lr0.005_bs10_epoch5_wd0.01 This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.8801 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.005 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| No log \| 1.0 \| 63 \| 3.3912 \| \| 3.5671 \| 2.0 \| 126 \| 3.4671 \| \| 3.5671 \| 3.0 \| 189 \| 3.6533 \| \| 3.3457 \| 4.0 \| 252 \| 3.7318 \| \| 3.1993 \| 5.0 \| 315 \| 3.8801 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
Edmon02/speecht5_finetuned_hy	Edmon02	2024-01-12T20:42:32Z	23	1	transformers	[ "transformers", "tensorboard", "safetensors", "speecht5", "text-to-audio", "generated_from_trainer", "text-to-speech", "hy", "en", "nl", "dataset:mozilla-foundation/common_voice_11_0", "base_model:microsoft/speecht5_tts", "base_model:finetune:microsoft/speecht5_tts", "license:mit", "region:us" ]	text-to-speech	2023-12-07T18:57:17Z	--- inference: false license: mit base_model: microsoft/speecht5_tts tags: - generated_from_trainer model-index: - name: speecht5_finetuned_hy results: [] language: - hy - en - nl datasets: - mozilla-foundation/common_voice_11_0 pipeline_tag: text-to-speech --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # speecht5_finetuned_hy This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.4785 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 1e-05 - train_batch_size: 4 - eval_batch_size: 2 - seed: 42 - gradient_accumulation_steps: 8 - total_train_batch_size: 32 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 125 - training_steps: 1000 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| 0.5547 \| 2.04 \| 250 \| 0.4983 \| \| 0.525 \| 4.07 \| 500 \| 0.4864 \| \| 0.52 \| 6.11 \| 750 \| 0.4812 \| \| 0.5286 \| 8.15 \| 1000 \| 0.4785 \| ### Framework versions - Transformers 4.35.2 - Pytorch 2.1.0+cu121 - Datasets 2.16.0 - Tokenizers 0.15.0
RiverTest/MTG3	RiverTest	2024-01-12T20:40:39Z	2	0	peft	[ "peft", "safetensors", "mistral", "trl", "sft", "generated_from_trainer", "dataset:generator", "base_model:teknium/OpenHermes-2.5-Mistral-7B", "base_model:adapter:teknium/OpenHermes-2.5-Mistral-7B", "license:apache-2.0", "region:us" ]	null	2024-01-12T20:29:48Z	--- license: apache-2.0 library_name: peft tags: - trl - sft - generated_from_trainer datasets: - generator base_model: teknium/OpenHermes-2.5-Mistral-7B model-index: - name: teknium/OpenHermes-2.5-Mistral-7B results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # teknium/OpenHermes-2.5-Mistral-7B This model is a fine-tuned version of [teknium/OpenHermes-2.5-Mistral-7B](https://huggingface.co/teknium/OpenHermes-2.5-Mistral-7B) on the generator dataset. It achieves the following results on the evaluation set: - Loss: 0.4764 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2.5e-05 - train_batch_size: 32 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 0.03 - training_steps: 220 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| 1.6429 \| 0.23 \| 10 \| 1.3400 \| \| 1.2684 \| 0.47 \| 20 \| 1.0591 \| \| 1.0218 \| 0.7 \| 30 \| 0.8600 \| \| 0.8095 \| 0.93 \| 40 \| 0.7113 \| \| 0.6887 \| 1.16 \| 50 \| 0.6353 \| \| 0.6207 \| 1.4 \| 60 \| 0.5958 \| \| 0.5726 \| 1.63 \| 70 \| 0.5717 \| \| 0.5639 \| 1.86 \| 80 \| 0.5548 \| \| 0.5496 \| 2.09 \| 90 \| 0.5409 \| \| 0.5353 \| 2.33 \| 100 \| 0.5297 \| \| 0.4968 \| 2.56 \| 110 \| 0.5204 \| \| 0.5031 \| 2.79 \| 120 \| 0.5118 \| \| 0.4906 \| 3.02 \| 130 \| 0.5053 \| \| 0.4671 \| 3.26 \| 140 \| 0.5015 \| \| 0.4734 \| 3.49 \| 150 \| 0.4960 \| \| 0.4734 \| 3.72 \| 160 \| 0.4890 \| \| 0.4522 \| 3.95 \| 170 \| 0.4857 \| \| 0.44 \| 4.19 \| 180 \| 0.4824 \| \| 0.4376 \| 4.42 \| 190 \| 0.4797 \| \| 0.4361 \| 4.65 \| 200 \| 0.4778 \| \| 0.4428 \| 4.88 \| 210 \| 0.4769 \| \| 0.4591 \| 5.12 \| 220 \| 0.4764 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.1.2+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF	TheBloke	2024-01-12T20:15:22Z	419	18	transformers	[ "transformers", "gguf", "mixtral", "base_model:NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss", "base_model:quantized:NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss", "license:cc-by-nc-4.0", "region:us", "conversational" ]	null	2024-01-12T18:32:48Z	--- base_model: NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss inference: false license: cc-by-nc-4.0 model_creator: NeverSleep model_name: Noromaid V0.4 Mixtral Instruct 8X7B ZLoss model_type: mixtral prompt_template: '<\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ' quantized_by: TheBloke --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Noromaid V0.4 Mixtral Instruct 8X7B ZLoss - GGUF - Model creator: [NeverSleep](https://huggingface.co/NeverSleep) - Original model: [Noromaid V0.4 Mixtral Instruct 8X7B ZLoss](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss) <!-- description start --> ## Description This repo contains GGUF format model files for [NeverSleep's Noromaid V0.4 Mixtral Instruct 8X7B ZLoss](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss). These files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/). <!-- description end --> <!-- README_GGUF.md-about-gguf start --> ### About GGUF GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp. Here is an incomplete list of clients and libraries that are known to support GGUF: * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option. * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration. * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling. * [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel. * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023. * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection. * [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration. * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server. * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use. * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models. <!-- README_GGUF.md-about-gguf end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF) * [NeverSleep's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: ChatML ``` <\|im_start\|>system {system_message}<\|im_end\|> <\|im_start\|>user {prompt}<\|im_end\|> <\|im_start\|>assistant ``` <!-- prompt-template end --> <!-- compatibility_gguf start --> ## Compatibility These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) They are also compatible with many third party UIs and libraries - please see the list at the top of this README. ## Explanation of quantisation methods <details> <summary>Click to see details</summary> The new methods available are: * GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw) * GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw. * GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw. * GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw * GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw Refer to the Provided Files table below to see what files use which methods, and how. </details> <!-- compatibility_gguf end --> <!-- README_GGUF.md-provided-files start --> ## Provided files \| Name \| Quant method \| Bits \| Size \| Max RAM required \| Use case \| \| ---- \| ---- \| ---- \| ---- \| ---- \| ----- \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q2_K.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q2_K.gguf) \| Q2_K \| 2 \| 17.17 GB\| 19.67 GB \| smallest, significant quality loss - not recommended for most purposes \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q3_K_M.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q3_K_M.gguf) \| Q3_K_M \| 3 \| 22.48 GB\| 24.98 GB \| very small, high quality loss \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_0.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_0.gguf) \| Q4_0 \| 4 \| 26.44 GB\| 28.94 GB \| legacy; small, very high quality loss - prefer using Q3_K_M \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf) \| Q4_K_M \| 4 \| 28.38 GB\| 30.88 GB \| medium, balanced quality - recommended \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q5_0.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q5_0.gguf) \| Q5_0 \| 5 \| 32.23 GB\| 34.73 GB \| legacy; medium, balanced quality - prefer using Q4_K_M \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q5_K_M.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q5_K_M.gguf) \| Q5_K_M \| 5 \| 33.23 GB\| 35.73 GB \| large, very low quality loss - recommended \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q6_K.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q6_K.gguf) \| Q6_K \| 6 \| 38.38 GB\| 40.88 GB \| very large, extremely low quality loss \| \| [noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q8_0.gguf](https://huggingface.co/TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF/blob/main/noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q8_0.gguf) \| Q8_0 \| 8 \| 49.62 GB\| 52.12 GB \| very large, extremely low quality loss - not recommended \| Note: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead. <!-- README_GGUF.md-provided-files end --> <!-- README_GGUF.md-how-to-download start --> ## How to download GGUF files Note for manual downloaders: You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file. The following clients/libraries will automatically download models for you, providing a list of available models to choose from: * LM Studio * LoLLMS Web UI * Faraday.dev ### In `text-generation-webui` Under Download Model, you can enter the model repo: TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF and below it, a specific filename to download, such as: noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf. Then click Download. ### On the command line, including multiple files at once I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` Then you can download any individual model file to the current directory, at high speed, with a command like this: ```shell huggingface-cli download TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage (click to read)</summary> You can also download multiple files at once with a pattern: ```shell huggingface-cli download TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF --local-dir . --local-dir-use-symlinks False --include='Q4_Kgguf' ``` For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> <!-- README_GGUF.md-how-to-download end --> <!-- README_GGUF.md-how-to-run start --> ## Example `llama.cpp` command Make sure you are using `llama.cpp` from commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later. ```shell ./main -ngl 35 -m noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf --color -c 32768 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "<\|im_start\|>system\n{system_message}<\|im_end\|>\n<\|im_start\|>user\n{prompt}<\|im_end\|>\n<\|im_start\|>assistant" ``` Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration. Change `-c 32768` to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically. Note that longer sequence lengths require much more resources, so you may need to reduce this value. If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins` For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md) ## How to run in `text-generation-webui` Further instructions can be found in the text-generation-webui documentation, here: [text-generation-webui/docs/04 ‐ Model Tab.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/04%20%E2%80%90%20Model%20Tab.md#llamacpp). ## How to run from Python code You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) or [ctransformers](https://github.com/marella/ctransformers) libraries. Note that at the time of writing (Nov 27th 2023), ctransformers has not been updated for some time and is not compatible with some recent models. Therefore I recommend you use llama-cpp-python. ### How to load this model in Python code, using llama-cpp-python For full documentation, please see: [llama-cpp-python docs](https://abetlen.github.io/llama-cpp-python/). #### First install the package Run one of the following commands, according to your system: ```shell # Base ctransformers with no GPU acceleration pip install llama-cpp-python # With NVidia CUDA acceleration CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install llama-cpp-python # Or with OpenBLAS acceleration CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python # Or with CLBLast acceleration CMAKE_ARGS="-DLLAMA_CLBLAST=on" pip install llama-cpp-python # Or with AMD ROCm GPU acceleration (Linux only) CMAKE_ARGS="-DLLAMA_HIPBLAS=on" pip install llama-cpp-python # Or with Metal GPU acceleration for macOS systems only CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python # In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA: $env:CMAKE_ARGS = "-DLLAMA_OPENBLAS=on" pip install llama-cpp-python ``` #### Simple llama-cpp-python example code ```python from llama_cpp import Llama # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system. llm = Llama( model_path="./noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf", # Download the model file first n_ctx=32768, # The max sequence length to use - note that longer sequence lengths require much more resources n_threads=8, # The number of CPU threads to use, tailor to your system and the resulting performance n_gpu_layers=35 # The number of layers to offload to GPU, if you have GPU acceleration available ) # Simple inference example output = llm( "<\|im_start\|>system\n{system_message}<\|im_end\|>\n<\|im_start\|>user\n{prompt}<\|im_end\|>\n<\|im_start\|>assistant", # Prompt max_tokens=512, # Generate up to 512 tokens stop=["</s>"], # Example stop token - not necessarily correct for this specific model! Please check before using. echo=True # Whether to echo the prompt ) # Chat Completion API llm = Llama(model_path="./noromaid-v0.4-mixtral-instruct-8x7b-zloss.Q4_K_M.gguf", chat_format="llama-2") # Set chat_format according to the model you are using llm.create_chat_completion( messages = [ {"role": "system", "content": "You are a story writing assistant."}, { "role": "user", "content": "Write a story about llamas." } ] ) ``` ## How to use with LangChain Here are guides on using llama-cpp-python and ctransformers with LangChain: * [LangChain + llama-cpp-python](https://python.langchain.com/docs/integrations/llms/llamacpp) * [LangChain + ctransformers](https://python.langchain.com/docs/integrations/providers/ctransformers) <!-- README_GGUF.md-how-to-run end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> <!-- original-model-card start --> # Original model card: NeverSleep's Noromaid V0.4 Mixtral Instruct 8X7B ZLoss ![image/png](https://cdn-uploads.huggingface.co/production/uploads/630dfb008df86f1e5becadc3/vwcJfOnL-2QDJ0ShfxRJ5.png) --- # Disclaimer: ## This model is experimental, do not expect everything to work. This model uses the Chatml prompting format --- Beeg noromaid on *steroids*. Suitable for RP, ERP. This model was trained on the Zloss fork of Charles, and should fix issue the model had. Use Chatml prompt format, but not the special token. The reason is that Axolotl merge the finetune with the base model at 1.0 weight basically, but this is too much, so I use another script available [HERE](https://github.com/DocShotgun/LLM-notebooks/blob/main/weighted-lora-merge.ipynb) to merge with less weight, sadly, it don't take the special Chatml token. It's like Orca2 for the matter. ## Credits: - Undi - IkariDev <!-- description start --> ## Description <!-- [Recommended settings - contributed by localfultonextractor](https://files.catbox.moe/ue0tja.json) --> This repo contains FP16 files of Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss. [FP16 - by IkariDev and Undi](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss) <!-- [GGUF - By TheBloke](https://huggingface.co/TheBloke/Athena-v4-GGUF)--> <!-- [GPTQ - By TheBloke](https://huggingface.co/TheBloke/Athena-v4-GPTQ)--> <!-- [exl2[8bpw-8h] - by AzureBlack](https://huggingface.co/AzureBlack/Echidna-13b-v0.3-8bpw-8h-exl2)--> <!-- [AWQ - By TheBloke](https://huggingface.co/TheBloke/Athena-v4-AWQ)--> <!-- [fp16 - by IkariDev+Undi95](https://huggingface.co/IkariDev/Athena-v4)--> [GGUF - by IkariDev and Undi](https://huggingface.co/NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss-GGUF) <!-- [OLD(GGUF - by IkariDev+Undi95)](https://huggingface.co/IkariDev/Athena-v4-GGUF)--> ## Ratings: Note: We have permission of all users to upload their ratings, we DONT screenshot random reviews without asking if we can put them here! No ratings yet! If you want your rating to be here, send us a message over on DC and we'll put up a screenshot of it here. DC name is "ikaridev" and "undi". <!-- description end --> <!-- prompt-template start --> ### Prompt format: Chatml ``` <\|im_start\|>system {sysprompt}<\|im_end\|> <\|im_start\|>user {input}<\|im_end\|> <\|im_start\|>assistant {output}<\|im_end\|> ``` ## Datasets used: - Aesir 1, 2 & 3 modified by us, credit to ([MinervaAI](https://huggingface.co/MinervaAI) / [Gryphe](https://huggingface.co/Gryphe)) - [LimaRP-20231109](https://huggingface.co/datasets/lemonilia/LimaRP) ([Lemonilia](https://huggingface.co/lemonilia)) - [ToxicQAFinal](https://huggingface.co/datasets/NobodyExistsOnTheInternet/ToxicQAFinal) ([NobodyExistsOnTheInternet](https://huggingface.co/NobodyExistsOnTheInternet) - [No-robots-ShareGPT](https://huggingface.co/datasets/Doctor-Shotgun/no-robots-sharegpt) ([Doctor-Shotgun](https://huggingface.co/Doctor-Shotgun)) ## Others Undi: If you want to support me, you can [here](https://ko-fi.com/undiai). IkariDev: Visit my [retro/neocities style website](https://ikaridevgit.github.io/) please kek <!-- original-model-card end -->
TheBloke/TenyxChat-7B-v1-GPTQ	TheBloke	2024-01-12T20:14:42Z	17	1	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "tenyx-fine-tuning", "dpo", "tenyxchat", "conversational", "en", "arxiv:2305.18290", "arxiv:2309.11235", "arxiv:2306.05685", "arxiv:1803.05457", "arxiv:1905.07830", "arxiv:2009.03300", "arxiv:2109.07958", "arxiv:1907.10641", "arxiv:2110.14168", "base_model:tenyx/TenyxChat-7B-v1", "base_model:quantized:tenyx/TenyxChat-7B-v1", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "4-bit", "gptq", "region:us" ]	text-generation	2024-01-12T19:46:35Z	--- base_model: tenyx/TenyxChat-7B-v1 inference: false language: - en library_name: transformers license: apache-2.0 model_creator: Tenyx model_name: TenyxChat 7B v1 model_type: mistral prompt_template: 'System: {system_message} User: {prompt} Assistant: ' quantized_by: TheBloke tags: - tenyx-fine-tuning - dpo - tenyxchat --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # TenyxChat 7B v1 - GPTQ - Model creator: [Tenyx](https://huggingface.co/tenyx) - Original model: [TenyxChat 7B v1](https://huggingface.co/tenyx/TenyxChat-7B-v1) <!-- description start --> # Description This repo contains GPTQ model files for [Tenyx's TenyxChat 7B v1](https://huggingface.co/tenyx/TenyxChat-7B-v1). Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them. These files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/). <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/TenyxChat-7B-v1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF) * [Tenyx's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/tenyx/TenyxChat-7B-v1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: System-User-Assistant-nohash ``` System: {system_message} User: {prompt} Assistant: ``` <!-- prompt-template end --> <!-- README_GPTQ.md-compatible clients start --> ## Known compatible clients / servers GPTQ models are currently supported on Linux (NVidia/AMD) and Windows (NVidia only). macOS users: please use GGUF models. These GPTQ models are known to work in the following inference servers/webuis. - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) - [KoboldAI United](https://github.com/henk717/koboldai) - [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui) - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) This may not be a complete list; if you know of others, please let me know! <!-- README_GPTQ.md-compatible clients end --> <!-- README_GPTQ.md-provided-files start --> ## Provided files, and GPTQ parameters Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements. Each separate quant is in a different branch. See below for instructions on fetching from different branches. Most GPTQ files are made with AutoGPTQ. Mistral models are currently made with Transformers. <details> <summary>Explanation of GPTQ parameters</summary> - Bits: The bit size of the quantised model. - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value. - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now. - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy. - GPTQ dataset: The calibration dataset used during quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ calibration dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s). - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences. - ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama and Mistral models in 4-bit. </details> \| Branch \| Bits \| GS \| Act Order \| Damp % \| GPTQ Dataset \| Seq Len \| Size \| ExLlama \| Desc \| \| ------ \| ---- \| -- \| --------- \| ------ \| ------------ \| ------- \| ---- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ/tree/main) \| 4 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.16 GB \| Yes \| 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. \| \| [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ/tree/gptq-4bit-32g-actorder_True) \| 4 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.57 GB \| Yes \| 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. \| \| [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ/tree/gptq-8bit--1g-actorder_True) \| 8 \| None \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 7.52 GB \| No \| 8-bit, with Act Order. No group size, to lower VRAM requirements. \| \| [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ/tree/gptq-8bit-128g-actorder_True) \| 8 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 7.68 GB \| No \| 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. \| \| [gptq-8bit-32g-actorder_True](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ/tree/gptq-8bit-32g-actorder_True) \| 8 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 8.17 GB \| No \| 8-bit, with group size 32g and Act Order for maximum inference quality. \| \| [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ/tree/gptq-4bit-64g-actorder_True) \| 4 \| 64 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.30 GB \| Yes \| 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. \| <!-- README_GPTQ.md-provided-files end --> <!-- README_GPTQ.md-download-from-branches start --> ## How to download, including from branches ### In text-generation-webui To download from the `main` branch, enter `TheBloke/TenyxChat-7B-v1-GPTQ` in the "Download model" box. To download from another branch, add `:branchname` to the end of the download name, eg `TheBloke/TenyxChat-7B-v1-GPTQ:gptq-4bit-32g-actorder_True` ### From the command line I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` To download the `main` branch to a folder called `TenyxChat-7B-v1-GPTQ`: ```shell mkdir TenyxChat-7B-v1-GPTQ huggingface-cli download TheBloke/TenyxChat-7B-v1-GPTQ --local-dir TenyxChat-7B-v1-GPTQ --local-dir-use-symlinks False ``` To download from a different branch, add the `--revision` parameter: ```shell mkdir TenyxChat-7B-v1-GPTQ huggingface-cli download TheBloke/TenyxChat-7B-v1-GPTQ --revision gptq-4bit-32g-actorder_True --local-dir TenyxChat-7B-v1-GPTQ --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage</summary> If you remove the `--local-dir-use-symlinks False` parameter, the files will instead be stored in the central Hugging Face cache directory (default location on Linux is: `~/.cache/huggingface`), and symlinks will be added to the specified `--local-dir`, pointing to their real location in the cache. This allows for interrupted downloads to be resumed, and allows you to quickly clone the repo to multiple places on disk without triggering a download again. The downside, and the reason why I don't list that as the default option, is that the files are then hidden away in a cache folder and it's harder to know where your disk space is being used, and to clear it up if/when you want to remove a download model. The cache location can be changed with the `HF_HOME` environment variable, and/or the `--cache-dir` parameter to `huggingface-cli`. For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell mkdir TenyxChat-7B-v1-GPTQ HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/TenyxChat-7B-v1-GPTQ --local-dir TenyxChat-7B-v1-GPTQ --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> ### With `git` (not recommended) To clone a specific branch with `git`, use a command like this: ```shell git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ ``` Note that using Git with HF repos is strongly discouraged. It will be much slower than using `huggingface-hub`, and will use twice as much disk space as it has to store the model files twice (it stores every byte both in the intended target folder, and again in the `.git` folder as a blob.) <!-- README_GPTQ.md-download-from-branches end --> <!-- README_GPTQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/TenyxChat-7B-v1-GPTQ`. - To download from a specific branch, enter for example `TheBloke/TenyxChat-7B-v1-GPTQ:gptq-4bit-32g-actorder_True` - see Provided Files above for the list of branches for each option. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `TenyxChat-7B-v1-GPTQ` 7. The model will automatically load, and is now ready for use! 8. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. - Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file `quantize_config.json`. 9. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_GPTQ.md-text-generation-webui end --> <!-- README_GPTQ.md-use-from-tgi start --> ## Serving this model from Text Generation Inference (TGI) It's recommended to use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/TenyxChat-7B-v1-GPTQ --port 3000 --quantize gptq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires huggingface-hub 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''System: {system_message} User: {prompt} Assistant: ''' client = InferenceClient(endpoint_url) response = client.text_generation( prompt_template, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(f"Model output: {response}") ``` <!-- README_GPTQ.md-use-from-tgi end --> <!-- README_GPTQ.md-use-from-python start --> ## Python code example: inference from this GPTQ model ### Install the necessary packages Requires: Transformers 4.33.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later. ```shell pip3 install --upgrade transformers optimum # If using PyTorch 2.1 + CUDA 12.x: pip3 install --upgrade auto-gptq # or, if using PyTorch 2.1 + CUDA 11.x: pip3 install --upgrade auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ ``` If you are using PyTorch 2.0, you will need to install AutoGPTQ from source. Likewise if you have problems with the pre-built wheels, you should try building from source: ```shell pip3 uninstall -y auto-gptq git clone https://github.com/PanQiWei/AutoGPTQ cd AutoGPTQ git checkout v0.5.1 pip3 install . ``` ### Example Python code ```python from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline model_name_or_path = "TheBloke/TenyxChat-7B-v1-GPTQ" # To use a different branch, change revision # For example: revision="gptq-4bit-32g-actorder_True" model = AutoModelForCausalLM.from_pretrained(model_name_or_path, device_map="auto", trust_remote_code=False, revision="main") tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True) prompt = "Write a story about llamas" system_message = "You are a story writing assistant" prompt_template=f'''System: {system_message} User: {prompt} Assistant: ''' print("\n\n* Generate:") input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda() output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512) print(tokenizer.decode(output[0])) # Inference can also be done using transformers' pipeline print("* Pipeline:") pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(pipe(prompt_template)[0]['generated_text']) ``` <!-- README_GPTQ.md-use-from-python end --> <!-- README_GPTQ.md-compatibility start --> ## Compatibility The files provided are tested to work with Transformers. For non-Mistral models, AutoGPTQ can also be used directly. [ExLlama](https://github.com/turboderp/exllama) is compatible with Llama architecture models (including Mistral, Yi, DeepSeek, SOLAR, etc) in 4-bit. Please see the Provided Files table above for per-file compatibility. For a list of clients/servers, please see "Known compatible clients / servers", above. <!-- README_GPTQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Tenyx's TenyxChat 7B v1 # TenyxChat: Language Model Alignment using Tenyx Fine-tuning Introducing TenyxChat, a series of ChatGPT-like models trained to function as useful assistants through preference tuning, using Tenyx's recently released advanced fine-tuning technology ([VentureBeat article](https://venturebeat.com/ai/tenyx-aims-to-fix-llms-catastrophic-forgetting-problem/)). Our first chat model in the series, TenyxChat-7B-v1, is trained using the [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290) framework on the open-source AI feedback dataset [UltraFeedback](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). We fine-tune [Openchat-3.5](https://arxiv.org/pdf/2309.11235.pdf) with our proprietary approach ([blog](https://www.tenyx.com/post/forgetting-and-toxicity-in-llms-a-deep-dive-on-fine-tuning-methods), [service](https://www.tenyx.com/fine-tuning)), which shows an increase in [MT-Bench](https://arxiv.org/abs/2306.05685), without a drop in performance of the model on other benchmarks. Our approach aims to mitigate forgetting in LLMs in a computationally efficient manner, thereby enabling continual fine-tuning capabilities without altering the pre-trained output distribution. TenyxChat-7B-v1 was trained using eight A100s (80GB) for two hours, with a training setup obtained from HuggingFaceH4 ([GitHub](https://github.com/huggingface/alignment-handbook)). # Model details - Model type: Fine-tuned 7B model for chat. - License: Apache 2.0 - Base model: OpenChat 3.5 ([https://huggingface.co/openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5)) - Demo: [spaces/tenyx/TenyxChat-7B-v1](https://huggingface.co/spaces/tenyx/TenyxChat-7B-v1) ## Usage Our model uses a simple chat template based on OpenChat 3.5. The chat template usage with a Hugging face generation example is shown below. ### Chat Template (Jinja) ```rust {{ bos_token }} {% for message in messages %} {% if message['role'] == 'user' %} {{ 'User:' + message['content'] + eos_token }} {% elif message['role'] == 'system' %} {{ 'System:' + message['content'] + eos_token }} {% elif message['role'] == 'assistant' %} {{ 'Assistant:' + message['content'] + eos_token }} {% endif %} {% if loop.last and add_generation_prompt %}\n{{ 'Assistant:' }}{% endif %}\n{% endfor %} ``` ### Hugging face Example ```python import torch from transformers import pipeline pipe = pipeline("text-generation", model="tenyx/TenyxChat-7B-v1", torch_dtype=torch.bfloat16, device_map="auto") messages = [ {"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate."}, {"role": "user", "content": "Hi. I would like to make a hotel booking."}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=512, do_sample=False) ``` ### Output ``` <s> System:You are a friendly chatbot who always responds in the style of a pirate.<\|end_of_turn\|> User:Hi. I would like to make a hotel booking.<\|end_of_turn\|> Assistant: Ahoy there me hearty! Arr, ye be lookin' fer a place to rest yer weary bones, eh? Well then, let's set sail on this grand adventure and find ye a swell place to stay! To begin, tell me the location ye be seekin' and the dates ye be lookin' to set sail. And don't ye worry, me matey, I'll be sure to find ye a place that'll make ye feel like a king or queen on land! ``` # Performance At the time of release (Jan 2024), TenyxChat-7B-v1 is the highest-ranked 7B chat model on the MT-Bench evaluation available for download and commercial use. We list here the benchmark results on several standard setups while comparing popular 7B models as baselines. ## MT-Bench MT-Bench is a benchmark made up of 80 high-quality multi-turn questions. These questions fall into eight categories: Writing, Roleplay, Reasoning, Math, Coding, Extraction, STEM, and Humanities. The chat models are rated using GPT-4 on a scale of 1 to 10, with higher values corresponding to better responses. \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| GPT-4* \| 8.95625 \| 9.02500 \| 8.990625 \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.75625 \| 8.103125 \| \| Starling-lm-7B-alpha \| 8.42500 \| 7.68750 \| 8.056250 \| \| OpenChat-3.5 \| 8.18125 \| 7.41250 \| 7.796875 \| \| GPT-3.5-turbo* \| 8.07500 \| 7.81250 \| 7.943750 \| \| OpenLLM Leader-7B** \| 8.05000 \| 7.61250 \| 7.831250 \| values reported on [lmsys](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) ChatBot Arena The [OpenLLM Leader](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) as of Jan 5, 2024 is the merge model available as [samir-fama/SamirGPT-v1](https://huggingface.co/samir-fama/SamirGPT-v1) ![hexplot.png](assets/hexplot.png) ### Comparison with additional Open LLM LeaderBoard models \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.756250 \| 8.103125 \| \| SamirGPT-v1 \| 8.05000 \| 7.612500 \| 7.831250 \| \| FernandoGPT-v1 \| 8.08125 \| 7.256250 \| 7.668750 \| \| Go-Bruins-v2 \| 8.13750 \| 7.150000 \| 7.643750 \| \| mistral_tv-neural-marconroni \| 7.76875 \| 6.987500 \| 7.378125 \| \| neuronovo-7B-v0.2 \| 7.73750 \| 6.662500 \| 7.200000 \| \| neural-chat-7b-v3-3 \| 7.39375 \| 5.881250 \| 6.637500 \| ## LM Evaluation - Open LLM Leaderboard We assess models on 7 benchmarks using the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). This setup is based of that used for [Open LLM Leaderboard.](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) - [AI2 Reasoning Challenge](https://arxiv.org/abs/1803.05457) (25-shot) - grade-school science questions. - [HellaSwag](https://arxiv.org/abs/1905.07830) (10-shot) - commonsense inference test. - [MMLU](https://arxiv.org/abs/2009.03300) (5-shot) - multitask accuracy test covering 57 tasks. - [TruthfulQA](https://arxiv.org/abs/2109.07958) (0-shot) - test measuring model's propensity to reproduce online falsehoods. - [Winogrande](https://arxiv.org/abs/1907.10641) (5-shot) - Winograd benchmark for commonsense reasoning. - [GSM8k](https://arxiv.org/abs/2110.14168) (5-shot) - grade school math word problems test. These benchmarks test reasoning and knowledge in various tasks in few-shot settings (higher scores are better). \| Model \| MMLU \| Winogrande \| GSM8k \| ARC \| HellaSwag \| TruthfulQA \| Average \| \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 63.6 \| 72.3 \| 69.0 \| 62.7 \| 66.6 \| 46.7 \| 63.48 \| \| Starling-7B-alpha \| 63.5 \| 72.1 \| 67.9 \| 61.1 \| 66.1 \| 42.1 \| 62.13 \| \| OpenChat-3.5 \| 63.6 \| 72.1 \| 68.2 \| 61.3 \| 65.2 \| 41.8 \| 62.03 \| \| Mistral-7B \| 62.4 \| 74.0 \| 38.1 \| 57.2 \| 62.8 \| 37.8 \| 55.38 \| \| OpenLLM Leader-7B \| 64.3 \| 78.7 \| 73.3 \| 66.6 \| 68.4 \| 58.5 \| 68.3 \| Note:* While the Open LLM Leaderboard indicates that these chat models perform less effectively compared to the leading 7B model, it's important to note that the leading model struggles in the multi-turn chat setting of MT-Bench (as demonstrated in our evaluation [above](#comparison-with-additional-open-llm-leaderboard-models)). In contrast, TenyxChat-7B-v1 demonstrates robustness against common fine-tuning challenges, such as catastrophic forgetting. This unique feature enables TenyxChat-7B-v1 to excel not only in chat benchmarks like MT-Bench, but also in a wider range of general reasoning benchmarks on the Open LLM Leaderboard. # Limitations TenyxChat-7B-v1, like other small-sized language models, has its own set of limitations. We haven’t fine-tuned the model explicitly to align with human safety preferences. Therefore, it is capable of producing undesirable outputs, particularly when adversarially prompted. From our observation, the model still tends to struggle with tasks that involve reasoning and math questions. In some instances, it might generate verbose or extraneous content. # License TenyxChat-7B-v1, similar to OpenChat 3.5, is distributed under the Apache License 2.0. # Citation If you use TenyxChat-7B for your research, cite us as ``` @misc{tenyxchat2024, title={TenyxChat: Language Model Alignment using Tenyx Fine-tuning}, author={Tenyx}, year={2024}, } ```
sessex/tabi_LoRA	sessex	2024-01-12T20:12:50Z	2	1	diffusers	[ "diffusers", "tensorboard", "stable-diffusion-xl", "stable-diffusion-xl-diffusers", "text-to-image", "lora", "template:sd-lora", "base_model:stabilityai/stable-diffusion-xl-base-1.0", "base_model:adapter:stabilityai/stable-diffusion-xl-base-1.0", "license:openrail++", "region:us" ]	text-to-image	2024-01-12T20:12:27Z	--- tags: - stable-diffusion-xl - stable-diffusion-xl-diffusers - text-to-image - diffusers - lora - template:sd-lora base_model: stabilityai/stable-diffusion-xl-base-1.0 instance_prompt: a photo of TOK shoe license: openrail++ --- # SDXL LoRA DreamBooth - sessex/tabi_LoRA <Gallery /> ## Model description These are sessex/tabi_LoRA LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0. The weights were trained using [DreamBooth](https://dreambooth.github.io/). LoRA for the text encoder was enabled: False. Special VAE used for training: madebyollin/sdxl-vae-fp16-fix. ## Trigger words You should use a photo of TOK shoe to trigger the image generation. ## Download model Weights for this model are available in Safetensors format. [Download](sessex/tabi_LoRA/tree/main) them in the Files & versions tab.
ricciobell/ricciobell	ricciobell	2024-01-12T20:12:36Z	0	0	null	[ "region:us" ]	null	2024-01-12T20:08:25Z	--- language: - it ha un naso enorme un bambino con le ali che viene tradito dalla ragazza
evenicole/google-play-sentiment-analysis	evenicole	2024-01-12T19:59:27Z	93	0	transformers	[ "transformers", "tensorboard", "safetensors", "bert", "text-classification", "generated_from_trainer", "base_model:neuralmind/bert-base-portuguese-cased", "base_model:finetune:neuralmind/bert-base-portuguese-cased", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2024-01-12T19:58:49Z	--- license: mit base_model: neuralmind/bert-base-portuguese-cased tags: - generated_from_trainer metrics: - accuracy model-index: - name: google-play-sentiment-analysis results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # google-play-sentiment-analysis This model is a fine-tuned version of [neuralmind/bert-base-portuguese-cased](https://huggingface.co/neuralmind/bert-base-portuguese-cased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.0466 - Accuracy: 0.67 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3.0 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Accuracy \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:--------:\| \| 0.7547 \| 1.0 \| 1250 \| 0.7947 \| 0.658 \| \| 0.6233 \| 2.0 \| 2500 \| 0.8057 \| 0.684 \| \| 0.4251 \| 3.0 \| 3750 \| 1.0466 \| 0.67 \| ### Framework versions - Transformers 4.35.2 - Pytorch 2.1.0+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
am-infoweb/rap_phase2_11jan_15i_v2	am-infoweb	2024-01-12T19:57:15Z	91	0	transformers	[ "transformers", "pytorch", "xlm-roberta", "question-answering", "generated_from_trainer", "base_model:FacebookAI/xlm-roberta-base", "base_model:finetune:FacebookAI/xlm-roberta-base", "license:mit", "endpoints_compatible", "region:us" ]	question-answering	2024-01-12T16:25:39Z	--- license: mit base_model: xlm-roberta-base tags: - generated_from_trainer model-index: - name: rap_phase2_11jan_15i_v2 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # rap_phase2_11jan_15i_v2 This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.0070 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 3 - eval_batch_size: 3 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 10 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:-----:\|:---------------:\| \| 0.1506 \| 1.0 \| 8580 \| 0.0892 \| \| 0.0691 \| 2.0 \| 17160 \| 0.0349 \| \| 0.0763 \| 3.0 \| 25740 \| 0.0216 \| \| 0.0355 \| 4.0 \| 34320 \| 0.0133 \| \| 0.0057 \| 5.0 \| 42900 \| 0.0074 \| \| 0.0167 \| 6.0 \| 51480 \| 0.0129 \| \| 0.0 \| 7.0 \| 60060 \| 0.0002 \| \| 0.0013 \| 8.0 \| 68640 \| 0.0047 \| \| 0.0126 \| 9.0 \| 77220 \| 0.0038 \| \| 0.001 \| 10.0 \| 85800 \| 0.0070 \| ### Framework versions - Transformers 4.32.0.dev0 - Pytorch 2.0.1+cu117 - Datasets 2.13.1 - Tokenizers 0.13.3
augmdc/bigbird_mental2019_1	augmdc	2024-01-12T19:51:03Z	88	0	transformers	[ "transformers", "pytorch", "big_bird", "multiple-choice", "generated_from_trainer", "base_model:augmdc/bigbird_race_1", "base_model:finetune:augmdc/bigbird_race_1", "license:apache-2.0", "endpoints_compatible", "region:us" ]	multiple-choice	2024-01-12T16:12:56Z	--- license: apache-2.0 base_model: augmdc/bigbird_race_1 tags: - generated_from_trainer metrics: - accuracy model-index: - name: bigbird_mental2019_1 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # bigbird_mental2019_1 This model is a fine-tuned version of [augmdc/bigbird_race_1](https://huggingface.co/augmdc/bigbird_race_1) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.1074 - Accuracy: 0.4983 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 2 - total_train_batch_size: 16 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 2 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Accuracy \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:--------:\| \| 1.3804 \| 1.0 \| 929 \| 1.2248 \| 0.4474 \| \| 1.1988 \| 2.0 \| 1858 \| 1.1074 \| 0.4983 \| ### Framework versions - Transformers 4.33.0 - Pytorch 2.0.0 - Datasets 2.1.0 - Tokenizers 0.13.3
LC008/RLTaxiDriver	LC008	2024-01-12T19:49:38Z	0	0	null	[ "Taxi-v3", "q-learning", "reinforcement-learning", "custom-implementation", "model-index", "region:us" ]	reinforcement-learning	2024-01-12T19:49:36Z	--- tags: - Taxi-v3 - q-learning - reinforcement-learning - custom-implementation model-index: - name: RLTaxiDriver results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: Taxi-v3 type: Taxi-v3 metrics: - type: mean_reward value: 7.50 +/- 2.72 name: mean_reward verified: false --- # Q-Learning Agent playing1 Taxi-v3 This is a trained model of a Q-Learning agent playing Taxi-v3 . ## Usage ```python model = load_from_hub(repo_id="LC008/RLTaxiDriver", filename="q-learning.pkl") # Don't forget to check if you need to add additional attributes (is_slippery=False etc) env = gym.make(model["env_id"]) ```
LC008/q-FrozenLake-v1-4x4-noSlippery	LC008	2024-01-12T19:47:13Z	0	0	null	[ "FrozenLake-v1-4x4-no_slippery", "q-learning", "reinforcement-learning", "custom-implementation", "model-index", "region:us" ]	reinforcement-learning	2024-01-12T19:47:11Z	--- tags: - FrozenLake-v1-4x4-no_slippery - q-learning - reinforcement-learning - custom-implementation model-index: - name: q-FrozenLake-v1-4x4-noSlippery results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: FrozenLake-v1-4x4-no_slippery type: FrozenLake-v1-4x4-no_slippery metrics: - type: mean_reward value: 1.00 +/- 0.00 name: mean_reward verified: false --- # Q-Learning Agent playing1 FrozenLake-v1 This is a trained model of a Q-Learning agent playing FrozenLake-v1 . ## Usage ```python model = load_from_hub(repo_id="LC008/q-FrozenLake-v1-4x4-noSlippery", filename="q-learning.pkl") # Don't forget to check if you need to add additional attributes (is_slippery=False etc) env = gym.make(model["env_id"]) ```
TheBloke/TenyxChat-7B-v1-GGUF	TheBloke	2024-01-12T19:46:59Z	173	10	transformers	[ "transformers", "gguf", "mistral", "tenyx-fine-tuning", "dpo", "tenyxchat", "en", "arxiv:2305.18290", "arxiv:2309.11235", "arxiv:2306.05685", "arxiv:1803.05457", "arxiv:1905.07830", "arxiv:2009.03300", "arxiv:2109.07958", "arxiv:1907.10641", "arxiv:2110.14168", "base_model:tenyx/TenyxChat-7B-v1", "base_model:quantized:tenyx/TenyxChat-7B-v1", "license:apache-2.0", "region:us", "conversational" ]	null	2024-01-12T11:50:15Z	--- base_model: tenyx/TenyxChat-7B-v1 inference: false language: - en library_name: transformers license: apache-2.0 model_creator: Tenyx model_name: TenyxChat 7B v1 model_type: mistral prompt_template: 'System: {system_message} User: {prompt} Assistant: ' quantized_by: TheBloke tags: - tenyx-fine-tuning - dpo - tenyxchat --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # TenyxChat 7B v1 - GGUF - Model creator: [Tenyx](https://huggingface.co/tenyx) - Original model: [TenyxChat 7B v1](https://huggingface.co/tenyx/TenyxChat-7B-v1) <!-- description start --> ## Description This repo contains GGUF format model files for [Tenyx's TenyxChat 7B v1](https://huggingface.co/tenyx/TenyxChat-7B-v1). These files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/). <!-- description end --> <!-- README_GGUF.md-about-gguf start --> ### About GGUF GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp. Here is an incomplete list of clients and libraries that are known to support GGUF: * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option. * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration. * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling. * [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel. * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023. * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection. * [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration. * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server. * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use. * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models. <!-- README_GGUF.md-about-gguf end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/TenyxChat-7B-v1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF) * [Tenyx's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/tenyx/TenyxChat-7B-v1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: System-User-Assistant-nohash ``` System: {system_message} User: {prompt} Assistant: ``` <!-- prompt-template end --> <!-- compatibility_gguf start --> ## Compatibility These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) They are also compatible with many third party UIs and libraries - please see the list at the top of this README. ## Explanation of quantisation methods <details> <summary>Click to see details</summary> The new methods available are: * GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw) * GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw. * GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw. * GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw * GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw Refer to the Provided Files table below to see what files use which methods, and how. </details> <!-- compatibility_gguf end --> <!-- README_GGUF.md-provided-files start --> ## Provided files \| Name \| Quant method \| Bits \| Size \| Max RAM required \| Use case \| \| ---- \| ---- \| ---- \| ---- \| ---- \| ----- \| \| [tenyxchat-7b-v1.Q2_K.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q2_K.gguf) \| Q2_K \| 2 \| 2.70 GB\| 5.20 GB \| smallest, significant quality loss - not recommended for most purposes \| \| [tenyxchat-7b-v1.Q3_K_S.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q3_K_S.gguf) \| Q3_K_S \| 3 \| 3.16 GB\| 5.66 GB \| very small, high quality loss \| \| [tenyxchat-7b-v1.Q3_K_M.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q3_K_M.gguf) \| Q3_K_M \| 3 \| 3.52 GB\| 6.02 GB \| very small, high quality loss \| \| [tenyxchat-7b-v1.Q3_K_L.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q3_K_L.gguf) \| Q3_K_L \| 3 \| 3.82 GB\| 6.32 GB \| small, substantial quality loss \| \| [tenyxchat-7b-v1.Q4_0.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q4_0.gguf) \| Q4_0 \| 4 \| 4.11 GB\| 6.61 GB \| legacy; small, very high quality loss - prefer using Q3_K_M \| \| [tenyxchat-7b-v1.Q4_K_S.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q4_K_S.gguf) \| Q4_K_S \| 4 \| 4.14 GB\| 6.64 GB \| small, greater quality loss \| \| [tenyxchat-7b-v1.Q4_K_M.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q4_K_M.gguf) \| Q4_K_M \| 4 \| 4.37 GB\| 6.87 GB \| medium, balanced quality - recommended \| \| [tenyxchat-7b-v1.Q5_0.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q5_0.gguf) \| Q5_0 \| 5 \| 5.00 GB\| 7.50 GB \| legacy; medium, balanced quality - prefer using Q4_K_M \| \| [tenyxchat-7b-v1.Q5_K_S.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q5_K_S.gguf) \| Q5_K_S \| 5 \| 5.00 GB\| 7.50 GB \| large, low quality loss - recommended \| \| [tenyxchat-7b-v1.Q5_K_M.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q5_K_M.gguf) \| Q5_K_M \| 5 \| 5.13 GB\| 7.63 GB \| large, very low quality loss - recommended \| \| [tenyxchat-7b-v1.Q6_K.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q6_K.gguf) \| Q6_K \| 6 \| 5.94 GB\| 8.44 GB \| very large, extremely low quality loss \| \| [tenyxchat-7b-v1.Q8_0.gguf](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF/blob/main/tenyxchat-7b-v1.Q8_0.gguf) \| Q8_0 \| 8 \| 7.70 GB\| 10.20 GB \| very large, extremely low quality loss - not recommended \| Note: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead. <!-- README_GGUF.md-provided-files end --> <!-- README_GGUF.md-how-to-download start --> ## How to download GGUF files Note for manual downloaders: You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file. The following clients/libraries will automatically download models for you, providing a list of available models to choose from: * LM Studio * LoLLMS Web UI * Faraday.dev ### In `text-generation-webui` Under Download Model, you can enter the model repo: TheBloke/TenyxChat-7B-v1-GGUF and below it, a specific filename to download, such as: tenyxchat-7b-v1.Q4_K_M.gguf. Then click Download. ### On the command line, including multiple files at once I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` Then you can download any individual model file to the current directory, at high speed, with a command like this: ```shell huggingface-cli download TheBloke/TenyxChat-7B-v1-GGUF tenyxchat-7b-v1.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage (click to read)</summary> You can also download multiple files at once with a pattern: ```shell huggingface-cli download TheBloke/TenyxChat-7B-v1-GGUF --local-dir . --local-dir-use-symlinks False --include='Q4_Kgguf' ``` For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/TenyxChat-7B-v1-GGUF tenyxchat-7b-v1.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> <!-- README_GGUF.md-how-to-download end --> <!-- README_GGUF.md-how-to-run start --> ## Example `llama.cpp` command Make sure you are using `llama.cpp` from commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later. ```shell ./main -ngl 35 -m tenyxchat-7b-v1.Q4_K_M.gguf --color -c 8192 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "System: {system_message}\nUser: {prompt}\nAssistant:" ``` Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration. Change `-c 8192` to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically. Note that longer sequence lengths require much more resources, so you may need to reduce this value. If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins` For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md) ## How to run in `text-generation-webui` Further instructions can be found in the text-generation-webui documentation, here: [text-generation-webui/docs/04 ‐ Model Tab.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/04%20%E2%80%90%20Model%20Tab.md#llamacpp). ## How to run from Python code You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) or [ctransformers](https://github.com/marella/ctransformers) libraries. Note that at the time of writing (Nov 27th 2023), ctransformers has not been updated for some time and is not compatible with some recent models. Therefore I recommend you use llama-cpp-python. ### How to load this model in Python code, using llama-cpp-python For full documentation, please see: [llama-cpp-python docs](https://abetlen.github.io/llama-cpp-python/). #### First install the package Run one of the following commands, according to your system: ```shell # Base ctransformers with no GPU acceleration pip install llama-cpp-python # With NVidia CUDA acceleration CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install llama-cpp-python # Or with OpenBLAS acceleration CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python # Or with CLBLast acceleration CMAKE_ARGS="-DLLAMA_CLBLAST=on" pip install llama-cpp-python # Or with AMD ROCm GPU acceleration (Linux only) CMAKE_ARGS="-DLLAMA_HIPBLAS=on" pip install llama-cpp-python # Or with Metal GPU acceleration for macOS systems only CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python # In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA: $env:CMAKE_ARGS = "-DLLAMA_OPENBLAS=on" pip install llama-cpp-python ``` #### Simple llama-cpp-python example code ```python from llama_cpp import Llama # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system. llm = Llama( model_path="./tenyxchat-7b-v1.Q4_K_M.gguf", # Download the model file first n_ctx=8192, # The max sequence length to use - note that longer sequence lengths require much more resources n_threads=8, # The number of CPU threads to use, tailor to your system and the resulting performance n_gpu_layers=35 # The number of layers to offload to GPU, if you have GPU acceleration available ) # Simple inference example output = llm( "System: {system_message}\nUser: {prompt}\nAssistant:", # Prompt max_tokens=512, # Generate up to 512 tokens stop=["</s>"], # Example stop token - not necessarily correct for this specific model! Please check before using. echo=True # Whether to echo the prompt ) # Chat Completion API llm = Llama(model_path="./tenyxchat-7b-v1.Q4_K_M.gguf", chat_format="llama-2") # Set chat_format according to the model you are using llm.create_chat_completion( messages = [ {"role": "system", "content": "You are a story writing assistant."}, { "role": "user", "content": "Write a story about llamas." } ] ) ``` ## How to use with LangChain Here are guides on using llama-cpp-python and ctransformers with LangChain: * [LangChain + llama-cpp-python](https://python.langchain.com/docs/integrations/llms/llamacpp) * [LangChain + ctransformers](https://python.langchain.com/docs/integrations/providers/ctransformers) <!-- README_GGUF.md-how-to-run end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> <!-- original-model-card start --> # Original model card: Tenyx's TenyxChat 7B v1 # TenyxChat: Language Model Alignment using Tenyx Fine-tuning Introducing TenyxChat, a series of ChatGPT-like models trained to function as useful assistants through preference tuning, using Tenyx's recently released advanced fine-tuning technology ([VentureBeat article](https://venturebeat.com/ai/tenyx-aims-to-fix-llms-catastrophic-forgetting-problem/)). Our first chat model in the series, TenyxChat-7B-v1, is trained using the [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290) framework on the open-source AI feedback dataset [UltraFeedback](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). We fine-tune [Openchat-3.5](https://arxiv.org/pdf/2309.11235.pdf) with our proprietary approach ([blog](https://www.tenyx.com/post/forgetting-and-toxicity-in-llms-a-deep-dive-on-fine-tuning-methods), [service](https://www.tenyx.com/fine-tuning)), which shows an increase in [MT-Bench](https://arxiv.org/abs/2306.05685), without a drop in performance of the model on other benchmarks. Our approach aims to mitigate forgetting in LLMs in a computationally efficient manner, thereby enabling continual fine-tuning capabilities without altering the pre-trained output distribution. TenyxChat-7B-v1 was trained using eight A100s (80GB) for two hours, with a training setup obtained from HuggingFaceH4 ([GitHub](https://github.com/huggingface/alignment-handbook)). # Model details - Model type: Fine-tuned 7B model for chat. - License: Apache 2.0 - Base model: OpenChat 3.5 ([https://huggingface.co/openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5)) - Demo: [spaces/tenyx/TenyxChat-7B-v1](https://huggingface.co/spaces/tenyx/TenyxChat-7B-v1) ## Usage Our model uses a simple chat template based on OpenChat 3.5. The chat template usage with a Hugging face generation example is shown below. ### Chat Template (Jinja) ```rust {{ bos_token }} {% for message in messages %} {% if message['role'] == 'user' %} {{ 'User:' + message['content'] + eos_token }} {% elif message['role'] == 'system' %} {{ 'System:' + message['content'] + eos_token }} {% elif message['role'] == 'assistant' %} {{ 'Assistant:' + message['content'] + eos_token }} {% endif %} {% if loop.last and add_generation_prompt %}\n{{ 'Assistant:' }}{% endif %}\n{% endfor %} ``` ### Hugging face Example ```python import torch from transformers import pipeline pipe = pipeline("text-generation", model="tenyx/TenyxChat-7B-v1", torch_dtype=torch.bfloat16, device_map="auto") messages = [ {"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate."}, {"role": "user", "content": "Hi. I would like to make a hotel booking."}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=512, do_sample=False) ``` ### Output ``` <s> System:You are a friendly chatbot who always responds in the style of a pirate.<\|end_of_turn\|> User:Hi. I would like to make a hotel booking.<\|end_of_turn\|> Assistant: Ahoy there me hearty! Arr, ye be lookin' fer a place to rest yer weary bones, eh? Well then, let's set sail on this grand adventure and find ye a swell place to stay! To begin, tell me the location ye be seekin' and the dates ye be lookin' to set sail. And don't ye worry, me matey, I'll be sure to find ye a place that'll make ye feel like a king or queen on land! ``` # Performance At the time of release (Jan 2024), TenyxChat-7B-v1 is the highest-ranked 7B chat model on the MT-Bench evaluation available for download and commercial use. We list here the benchmark results on several standard setups while comparing popular 7B models as baselines. ## MT-Bench MT-Bench is a benchmark made up of 80 high-quality multi-turn questions. These questions fall into eight categories: Writing, Roleplay, Reasoning, Math, Coding, Extraction, STEM, and Humanities. The chat models are rated using GPT-4 on a scale of 1 to 10, with higher values corresponding to better responses. \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| GPT-4* \| 8.95625 \| 9.02500 \| 8.990625 \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.75625 \| 8.103125 \| \| Starling-lm-7B-alpha \| 8.42500 \| 7.68750 \| 8.056250 \| \| OpenChat-3.5 \| 8.18125 \| 7.41250 \| 7.796875 \| \| GPT-3.5-turbo* \| 8.07500 \| 7.81250 \| 7.943750 \| \| OpenLLM Leader-7B** \| 8.05000 \| 7.61250 \| 7.831250 \| values reported on [lmsys](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) ChatBot Arena The [OpenLLM Leader](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) as of Jan 5, 2024 is the merge model available as [samir-fama/SamirGPT-v1](https://huggingface.co/samir-fama/SamirGPT-v1) ![hexplot.png](assets/hexplot.png) ### Comparison with additional Open LLM LeaderBoard models \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.756250 \| 8.103125 \| \| SamirGPT-v1 \| 8.05000 \| 7.612500 \| 7.831250 \| \| FernandoGPT-v1 \| 8.08125 \| 7.256250 \| 7.668750 \| \| Go-Bruins-v2 \| 8.13750 \| 7.150000 \| 7.643750 \| \| mistral_tv-neural-marconroni \| 7.76875 \| 6.987500 \| 7.378125 \| \| neuronovo-7B-v0.2 \| 7.73750 \| 6.662500 \| 7.200000 \| \| neural-chat-7b-v3-3 \| 7.39375 \| 5.881250 \| 6.637500 \| ## LM Evaluation - Open LLM Leaderboard We assess models on 7 benchmarks using the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). This setup is based of that used for [Open LLM Leaderboard.](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) - [AI2 Reasoning Challenge](https://arxiv.org/abs/1803.05457) (25-shot) - grade-school science questions. - [HellaSwag](https://arxiv.org/abs/1905.07830) (10-shot) - commonsense inference test. - [MMLU](https://arxiv.org/abs/2009.03300) (5-shot) - multitask accuracy test covering 57 tasks. - [TruthfulQA](https://arxiv.org/abs/2109.07958) (0-shot) - test measuring model's propensity to reproduce online falsehoods. - [Winogrande](https://arxiv.org/abs/1907.10641) (5-shot) - Winograd benchmark for commonsense reasoning. - [GSM8k](https://arxiv.org/abs/2110.14168) (5-shot) - grade school math word problems test. These benchmarks test reasoning and knowledge in various tasks in few-shot settings (higher scores are better). \| Model \| MMLU \| Winogrande \| GSM8k \| ARC \| HellaSwag \| TruthfulQA \| Average \| \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 63.6 \| 72.3 \| 69.0 \| 62.7 \| 66.6 \| 46.7 \| 63.48 \| \| Starling-7B-alpha \| 63.5 \| 72.1 \| 67.9 \| 61.1 \| 66.1 \| 42.1 \| 62.13 \| \| OpenChat-3.5 \| 63.6 \| 72.1 \| 68.2 \| 61.3 \| 65.2 \| 41.8 \| 62.03 \| \| Mistral-7B \| 62.4 \| 74.0 \| 38.1 \| 57.2 \| 62.8 \| 37.8 \| 55.38 \| \| OpenLLM Leader-7B \| 64.3 \| 78.7 \| 73.3 \| 66.6 \| 68.4 \| 58.5 \| 68.3 \| Note:* While the Open LLM Leaderboard indicates that these chat models perform less effectively compared to the leading 7B model, it's important to note that the leading model struggles in the multi-turn chat setting of MT-Bench (as demonstrated in our evaluation [above](#comparison-with-additional-open-llm-leaderboard-models)). In contrast, TenyxChat-7B-v1 demonstrates robustness against common fine-tuning challenges, such as catastrophic forgetting. This unique feature enables TenyxChat-7B-v1 to excel not only in chat benchmarks like MT-Bench, but also in a wider range of general reasoning benchmarks on the Open LLM Leaderboard. # Limitations TenyxChat-7B-v1, like other small-sized language models, has its own set of limitations. We haven’t fine-tuned the model explicitly to align with human safety preferences. Therefore, it is capable of producing undesirable outputs, particularly when adversarially prompted. From our observation, the model still tends to struggle with tasks that involve reasoning and math questions. In some instances, it might generate verbose or extraneous content. # License TenyxChat-7B-v1, similar to OpenChat 3.5, is distributed under the Apache License 2.0. # Citation If you use TenyxChat-7B for your research, cite us as ``` @misc{tenyxchat2024, title={TenyxChat: Language Model Alignment using Tenyx Fine-tuning}, author={Tenyx}, year={2024}, } ``` <!-- original-model-card end -->
TheBloke/TenyxChat-7B-v1-AWQ	TheBloke	2024-01-12T19:46:49Z	68	4	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "tenyx-fine-tuning", "dpo", "tenyxchat", "conversational", "en", "arxiv:2305.18290", "arxiv:2309.11235", "arxiv:2306.05685", "arxiv:1803.05457", "arxiv:1905.07830", "arxiv:2009.03300", "arxiv:2109.07958", "arxiv:1907.10641", "arxiv:2110.14168", "base_model:tenyx/TenyxChat-7B-v1", "base_model:quantized:tenyx/TenyxChat-7B-v1", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "4-bit", "awq", "region:us" ]	text-generation	2024-01-12T11:50:14Z	--- base_model: tenyx/TenyxChat-7B-v1 inference: false language: - en library_name: transformers license: apache-2.0 model_creator: Tenyx model_name: TenyxChat 7B v1 model_type: mistral prompt_template: 'System: {system_message} User: {prompt} Assistant: ' quantized_by: TheBloke tags: - tenyx-fine-tuning - dpo - tenyxchat --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # TenyxChat 7B v1 - AWQ - Model creator: [Tenyx](https://huggingface.co/tenyx) - Original model: [TenyxChat 7B v1](https://huggingface.co/tenyx/TenyxChat-7B-v1) <!-- description start --> ## Description This repo contains AWQ model files for [Tenyx's TenyxChat 7B v1](https://huggingface.co/tenyx/TenyxChat-7B-v1). These files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/). ### About AWQ AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference with equivalent or better quality compared to the most commonly used GPTQ settings. AWQ models are currently supported on Linux and Windows, with NVidia GPUs only. macOS users: please use GGUF models instead. It is supported by: - [Text Generation Webui](https://github.com/oobabooga/text-generation-webui) - using Loader: AutoAWQ - [vLLM](https://github.com/vllm-project/vllm) - version 0.2.2 or later for support for all model types. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later, from any code or client that supports Transformers - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) - for use from Python code <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/TenyxChat-7B-v1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/TenyxChat-7B-v1-GGUF) * [Tenyx's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/tenyx/TenyxChat-7B-v1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: System-User-Assistant-nohash ``` System: {system_message} User: {prompt} Assistant: ``` <!-- prompt-template end --> <!-- README_AWQ.md-provided-files start --> ## Provided files, and AWQ parameters I currently release 128g GEMM models only. The addition of group_size 32 models, and GEMV kernel models, is being actively considered. Models are released as sharded safetensors files. \| Branch \| Bits \| GS \| AWQ Dataset \| Seq Len \| Size \| \| ------ \| ---- \| -- \| ----------- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/TenyxChat-7B-v1-AWQ/tree/main) \| 4 \| 128 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.15 GB <!-- README_AWQ.md-provided-files end --> <!-- README_AWQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/TenyxChat-7B-v1-AWQ`. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `TenyxChat-7B-v1-AWQ` 7. Select Loader: AutoAWQ. 8. Click Load, and the model will load and is now ready for use. 9. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. 10. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_AWQ.md-text-generation-webui end --> <!-- README_AWQ.md-use-from-vllm start --> ## Multi-user inference server: vLLM Documentation on installing and using vLLM [can be found here](https://vllm.readthedocs.io/en/latest/). - Please ensure you are using vLLM version 0.2 or later. - When using vLLM as a server, pass the `--quantization awq` parameter. For example: ```shell python3 -m vllm.entrypoints.api_server --model TheBloke/TenyxChat-7B-v1-AWQ --quantization awq --dtype auto ``` - When using vLLM from Python code, again set `quantization=awq`. For example: ```python from vllm import LLM, SamplingParams prompts = [ "Tell me about AI", "Write a story about llamas", "What is 291 - 150?", "How much wood would a woodchuck chuck if a woodchuck could chuck wood?", ] prompt_template=f'''System: {system_message} User: {prompt} Assistant: ''' prompts = [prompt_template.format(prompt=prompt) for prompt in prompts] sampling_params = SamplingParams(temperature=0.8, top_p=0.95) llm = LLM(model="TheBloke/TenyxChat-7B-v1-AWQ", quantization="awq", dtype="auto") outputs = llm.generate(prompts, sampling_params) # Print the outputs. for output in outputs: prompt = output.prompt generated_text = output.outputs[0].text print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}") ``` <!-- README_AWQ.md-use-from-vllm start --> <!-- README_AWQ.md-use-from-tgi start --> ## Multi-user inference server: Hugging Face Text Generation Inference (TGI) Use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/TenyxChat-7B-v1-AWQ --port 3000 --quantize awq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires [huggingface-hub](https://github.com/huggingface/huggingface_hub) 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''System: {system_message} User: {prompt} Assistant: ''' client = InferenceClient(endpoint_url) response = client.text_generation(prompt, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1) print(f"Model output: ", response) ``` <!-- README_AWQ.md-use-from-tgi end --> <!-- README_AWQ.md-use-from-python start --> ## Inference from Python code using Transformers ### Install the necessary packages - Requires: [Transformers](https://huggingface.co/docs/transformers) 4.35.0 or later. - Requires: [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) 0.1.6 or later. ```shell pip3 install --upgrade "autoawq>=0.1.6" "transformers>=4.35.0" ``` Note that if you are using PyTorch 2.0.1, the above AutoAWQ command will automatically upgrade you to PyTorch 2.1.0. If you are using CUDA 11.8 and wish to continue using PyTorch 2.0.1, instead run this command: ```shell pip3 install https://github.com/casper-hansen/AutoAWQ/releases/download/v0.1.6/autoawq-0.1.6+cu118-cp310-cp310-linux_x86_64.whl ``` If you have problems installing [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) using the pre-built wheels, install it from source instead: ```shell pip3 uninstall -y autoawq git clone https://github.com/casper-hansen/AutoAWQ cd AutoAWQ pip3 install . ``` ### Transformers example code (requires Transformers 4.35.0 and later) ```python from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer model_name_or_path = "TheBloke/TenyxChat-7B-v1-AWQ" tokenizer = AutoTokenizer.from_pretrained(model_name_or_path) model = AutoModelForCausalLM.from_pretrained( model_name_or_path, low_cpu_mem_usage=True, device_map="cuda:0" ) # Using the text streamer to stream output one token at a time streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True) prompt = "Tell me about AI" prompt_template=f'''System: {system_message} User: {prompt} Assistant: ''' # Convert prompt to tokens tokens = tokenizer( prompt_template, return_tensors='pt' ).input_ids.cuda() generation_params = { "do_sample": True, "temperature": 0.7, "top_p": 0.95, "top_k": 40, "max_new_tokens": 512, "repetition_penalty": 1.1 } # Generate streamed output, visible one token at a time generation_output = model.generate( tokens, streamer=streamer, generation_params ) # Generation without a streamer, which will include the prompt in the output generation_output = model.generate( tokens, generation_params ) # Get the tokens from the output, decode them, print them token_output = generation_output[0] text_output = tokenizer.decode(token_output) print("model.generate output: ", text_output) # Inference is also possible via Transformers' pipeline from transformers import pipeline pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, *generation_params ) pipe_output = pipe(prompt_template)[0]['generated_text'] print("pipeline output: ", pipe_output) ``` <!-- README_AWQ.md-use-from-python end --> <!-- README_AWQ.md-compatibility start --> ## Compatibility The files provided are tested to work with: - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) using `Loader: AutoAWQ`. - [vLLM](https://github.com/vllm-project/vllm) version 0.2.0 and later. - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) version 1.1.0 and later. - [Transformers](https://huggingface.co/docs/transformers) version 4.35.0 and later. - [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) version 0.1.1 and later. <!-- README_AWQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Tenyx's TenyxChat 7B v1 # TenyxChat: Language Model Alignment using Tenyx Fine-tuning Introducing TenyxChat, a series of ChatGPT-like models trained to function as useful assistants through preference tuning, using Tenyx's recently released advanced fine-tuning technology ([VentureBeat article](https://venturebeat.com/ai/tenyx-aims-to-fix-llms-catastrophic-forgetting-problem/)). Our first chat model in the series, TenyxChat-7B-v1, is trained using the [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290) framework on the open-source AI feedback dataset [UltraFeedback](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). We fine-tune [Openchat-3.5](https://arxiv.org/pdf/2309.11235.pdf) with our proprietary approach ([blog](https://www.tenyx.com/post/forgetting-and-toxicity-in-llms-a-deep-dive-on-fine-tuning-methods), [service](https://www.tenyx.com/fine-tuning)), which shows an increase in [MT-Bench](https://arxiv.org/abs/2306.05685), without a drop in performance of the model on other benchmarks. Our approach aims to mitigate forgetting in LLMs in a computationally efficient manner, thereby enabling continual fine-tuning capabilities without altering the pre-trained output distribution. TenyxChat-7B-v1 was trained using eight A100s (80GB) for two hours, with a training setup obtained from HuggingFaceH4 ([GitHub](https://github.com/huggingface/alignment-handbook)). # Model details - Model type: Fine-tuned 7B model for chat. - License: Apache 2.0 - Base model: OpenChat 3.5 ([https://huggingface.co/openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5)) - Demo: [spaces/tenyx/TenyxChat-7B-v1](https://huggingface.co/spaces/tenyx/TenyxChat-7B-v1) ## Usage Our model uses a simple chat template based on OpenChat 3.5. The chat template usage with a Hugging face generation example is shown below. ### Chat Template (Jinja) ```rust {{ bos_token }} {% for message in messages %} {% if message['role'] == 'user' %} {{ 'User:' + message['content'] + eos_token }} {% elif message['role'] == 'system' %} {{ 'System:' + message['content'] + eos_token }} {% elif message['role'] == 'assistant' %} {{ 'Assistant:' + message['content'] + eos_token }} {% endif %} {% if loop.last and add_generation_prompt %}\n{{ 'Assistant:' }}{% endif %}\n{% endfor %} ``` ### Hugging face Example ```python import torch from transformers import pipeline pipe = pipeline("text-generation", model="tenyx/TenyxChat-7B-v1", torch_dtype=torch.bfloat16, device_map="auto") messages = [ {"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate."}, {"role": "user", "content": "Hi. I would like to make a hotel booking."}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=512, do_sample=False) ``` ### Output ``` <s> System:You are a friendly chatbot who always responds in the style of a pirate.<\|end_of_turn\|> User:Hi. I would like to make a hotel booking.<\|end_of_turn\|> Assistant: Ahoy there me hearty! Arr, ye be lookin' fer a place to rest yer weary bones, eh? Well then, let's set sail on this grand adventure and find ye a swell place to stay! To begin, tell me the location ye be seekin' and the dates ye be lookin' to set sail. And don't ye worry, me matey, I'll be sure to find ye a place that'll make ye feel like a king or queen on land! ``` # Performance At the time of release (Jan 2024), TenyxChat-7B-v1 is the highest-ranked 7B chat model on the MT-Bench evaluation available for download and commercial use. We list here the benchmark results on several standard setups while comparing popular 7B models as baselines. ## MT-Bench MT-Bench is a benchmark made up of 80 high-quality multi-turn questions. These questions fall into eight categories: Writing, Roleplay, Reasoning, Math, Coding, Extraction, STEM, and Humanities. The chat models are rated using GPT-4 on a scale of 1 to 10, with higher values corresponding to better responses. \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| GPT-4* \| 8.95625 \| 9.02500 \| 8.990625 \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.75625 \| 8.103125 \| \| Starling-lm-7B-alpha \| 8.42500 \| 7.68750 \| 8.056250 \| \| OpenChat-3.5 \| 8.18125 \| 7.41250 \| 7.796875 \| \| GPT-3.5-turbo* \| 8.07500 \| 7.81250 \| 7.943750 \| \| OpenLLM Leader-7B** \| 8.05000 \| 7.61250 \| 7.831250 \| values reported on [lmsys](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) ChatBot Arena The [OpenLLM Leader](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) as of Jan 5, 2024 is the merge model available as [samir-fama/SamirGPT-v1](https://huggingface.co/samir-fama/SamirGPT-v1) ![hexplot.png](assets/hexplot.png) ### Comparison with additional Open LLM LeaderBoard models \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.756250 \| 8.103125 \| \| SamirGPT-v1 \| 8.05000 \| 7.612500 \| 7.831250 \| \| FernandoGPT-v1 \| 8.08125 \| 7.256250 \| 7.668750 \| \| Go-Bruins-v2 \| 8.13750 \| 7.150000 \| 7.643750 \| \| mistral_tv-neural-marconroni \| 7.76875 \| 6.987500 \| 7.378125 \| \| neuronovo-7B-v0.2 \| 7.73750 \| 6.662500 \| 7.200000 \| \| neural-chat-7b-v3-3 \| 7.39375 \| 5.881250 \| 6.637500 \| ## LM Evaluation - Open LLM Leaderboard We assess models on 7 benchmarks using the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). This setup is based of that used for [Open LLM Leaderboard.](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) - [AI2 Reasoning Challenge](https://arxiv.org/abs/1803.05457) (25-shot) - grade-school science questions. - [HellaSwag](https://arxiv.org/abs/1905.07830) (10-shot) - commonsense inference test. - [MMLU](https://arxiv.org/abs/2009.03300) (5-shot) - multitask accuracy test covering 57 tasks. - [TruthfulQA](https://arxiv.org/abs/2109.07958) (0-shot) - test measuring model's propensity to reproduce online falsehoods. - [Winogrande](https://arxiv.org/abs/1907.10641) (5-shot) - Winograd benchmark for commonsense reasoning. - [GSM8k](https://arxiv.org/abs/2110.14168) (5-shot) - grade school math word problems test. These benchmarks test reasoning and knowledge in various tasks in few-shot settings (higher scores are better). \| Model \| MMLU \| Winogrande \| GSM8k \| ARC \| HellaSwag \| TruthfulQA \| Average \| \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 63.6 \| 72.3 \| 69.0 \| 62.7 \| 66.6 \| 46.7 \| 63.48 \| \| Starling-7B-alpha \| 63.5 \| 72.1 \| 67.9 \| 61.1 \| 66.1 \| 42.1 \| 62.13 \| \| OpenChat-3.5 \| 63.6 \| 72.1 \| 68.2 \| 61.3 \| 65.2 \| 41.8 \| 62.03 \| \| Mistral-7B \| 62.4 \| 74.0 \| 38.1 \| 57.2 \| 62.8 \| 37.8 \| 55.38 \| \| OpenLLM Leader-7B \| 64.3 \| 78.7 \| 73.3 \| 66.6 \| 68.4 \| 58.5 \| 68.3 \| Note:* While the Open LLM Leaderboard indicates that these chat models perform less effectively compared to the leading 7B model, it's important to note that the leading model struggles in the multi-turn chat setting of MT-Bench (as demonstrated in our evaluation [above](#comparison-with-additional-open-llm-leaderboard-models)). In contrast, TenyxChat-7B-v1 demonstrates robustness against common fine-tuning challenges, such as catastrophic forgetting. This unique feature enables TenyxChat-7B-v1 to excel not only in chat benchmarks like MT-Bench, but also in a wider range of general reasoning benchmarks on the Open LLM Leaderboard. # Limitations TenyxChat-7B-v1, like other small-sized language models, has its own set of limitations. We haven’t fine-tuned the model explicitly to align with human safety preferences. Therefore, it is capable of producing undesirable outputs, particularly when adversarially prompted. From our observation, the model still tends to struggle with tasks that involve reasoning and math questions. In some instances, it might generate verbose or extraneous content. # License TenyxChat-7B-v1, similar to OpenChat 3.5, is distributed under the Apache License 2.0. # Citation If you use TenyxChat-7B for your research, cite us as ``` @misc{tenyxchat2024, title={TenyxChat: Language Model Alignment using Tenyx Fine-tuning}, author={Tenyx}, year={2024}, } ```
mitro99/distilhubert-finetuned-gtzan_batch8_grad4_cosine	mitro99	2024-01-12T19:28:42Z	145	0	transformers	[ "transformers", "tensorboard", "safetensors", "hubert", "audio-classification", "generated_from_trainer", "dataset:marsyas/gtzan", "base_model:ntu-spml/distilhubert", "base_model:finetune:ntu-spml/distilhubert", "license:apache-2.0", "model-index", "endpoints_compatible", "region:us" ]	audio-classification	2024-01-12T19:28:15Z	--- license: apache-2.0 base_model: ntu-spml/distilhubert tags: - generated_from_trainer datasets: - marsyas/gtzan metrics: - accuracy model-index: - name: distilhubert-finetuned-gtzan results: - task: name: Audio Classification type: audio-classification dataset: name: GTZAN type: marsyas/gtzan config: all split: train args: all metrics: - name: Accuracy type: accuracy value: 0.88 --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilhubert-finetuned-gtzan This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset. It achieves the following results on the evaluation set: - Accuracy: 0.88 - Loss: 0.4331 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 32 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 20 - mixed_precision_training: Native AMP ### Training results \| Training Loss \| Epoch \| Step \| Accuracy \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:--------:\|:---------------:\| \| 2.2693 \| 0.99 \| 28 \| 0.31 \| 2.2480 \| \| 1.9782 \| 1.98 \| 56 \| 0.45 \| 1.8990 \| \| 1.6438 \| 2.97 \| 84 \| 0.62 \| 1.5180 \| \| 1.3307 \| 4.0 \| 113 \| 0.73 \| 1.2206 \| \| 1.133 \| 4.99 \| 141 \| 0.76 \| 0.9961 \| \| 0.9384 \| 5.98 \| 169 \| 0.78 \| 0.8889 \| \| 0.8668 \| 6.97 \| 197 \| 0.79 \| 0.7543 \| \| 0.674 \| 8.0 \| 226 \| 0.79 \| 0.7433 \| \| 0.5997 \| 8.99 \| 254 \| 0.83 \| 0.6194 \| \| 0.5195 \| 9.98 \| 282 \| 0.91 \| 0.5685 \| \| 0.401 \| 10.97 \| 310 \| 0.91 \| 0.5144 \| \| 0.3151 \| 12.0 \| 339 \| 0.87 \| 0.4775 \| \| 0.2653 \| 12.99 \| 367 \| 0.88 \| 0.4984 \| \| 0.2182 \| 13.98 \| 395 \| 0.88 \| 0.4337 \| \| 0.2036 \| 14.97 \| 423 \| 0.89 \| 0.4657 \| \| 0.1925 \| 16.0 \| 452 \| 0.89 \| 0.4222 \| \| 0.1807 \| 16.99 \| 480 \| 0.87 \| 0.4512 \| \| 0.1626 \| 17.98 \| 508 \| 0.88 \| 0.4247 \| \| 0.1388 \| 18.97 \| 536 \| 0.88 \| 0.4324 \| \| 0.1718 \| 19.82 \| 560 \| 0.88 \| 0.4331 \| ### Framework versions - Transformers 4.36.2 - Pytorch 2.1.1+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0
foobb/opt-6.7b-lora	foobb	2024-01-12T19:22:34Z	0	0	peft	[ "peft", "safetensors", "arxiv:1910.09700", "base_model:facebook/opt-6.7b", "base_model:adapter:facebook/opt-6.7b", "region:us" ]	null	2024-01-12T19:22:30Z	--- library_name: peft base_model: facebook/opt-6.7b --- # Model Card for Model ID <!-- Provide a quick summary of what the model is/does. --> ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> - Developed by: [More Information Needed] - Funded by [optional]: [More Information Needed] - Shared by [optional]: [More Information Needed] - Model type: [More Information Needed] - Language(s) (NLP): [More Information Needed] - License: [More Information Needed] - Finetuned from model [optional]: [More Information Needed] ### Model Sources [optional] <!-- Provide the basic links for the model. --> - Repository: [More Information Needed] - Paper [optional]: [More Information Needed] - Demo [optional]: [More Information Needed] ## Uses <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> [More Information Needed] ### Downstream Use [optional] <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app --> [More Information Needed] ### Out-of-Scope Use <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> [More Information Needed] ## Bias, Risks, and Limitations <!-- This section is meant to convey both technical and sociotechnical limitations. --> [More Information Needed] ### Recommendations <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. --> Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. ## How to Get Started with the Model Use the code below to get started with the model. [More Information Needed] ## Training Details ### Training Data <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> [More Information Needed] ### Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> #### Preprocessing [optional] [More Information Needed] #### Training Hyperparameters - Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision --> #### Speeds, Sizes, Times [optional] <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. --> [More Information Needed] ## Evaluation <!-- This section describes the evaluation protocols and provides the results. --> ### Testing Data, Factors & Metrics #### Testing Data <!-- This should link to a Dataset Card if possible. --> [More Information Needed] #### Factors <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. --> [More Information Needed] #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> [More Information Needed] ### Results [More Information Needed] #### Summary ## Model Examination [optional] <!-- Relevant interpretability work for the model goes here --> [More Information Needed] ## Environmental Impact <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly --> Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - Hardware Type: [More Information Needed] - Hours used: [More Information Needed] - Cloud Provider: [More Information Needed] - Compute Region: [More Information Needed] - Carbon Emitted: [More Information Needed] ## Technical Specifications [optional] ### Model Architecture and Objective [More Information Needed] ### Compute Infrastructure [More Information Needed] #### Hardware [More Information Needed] #### Software [More Information Needed] ## Citation [optional] <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> BibTeX: [More Information Needed] APA: [More Information Needed] ## Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> [More Information Needed] ## More Information [optional] [More Information Needed] ## Model Card Authors [optional] [More Information Needed] ## Model Card Contact [More Information Needed] ### Framework versions - PEFT 0.7.2.dev0
jysssacc/mt0-base_fine_lr5e-05_bs1_epoch5_wd0.01	jysssacc	2024-01-12T19:21:18Z	89	0	transformers	[ "transformers", "safetensors", "mt5", "text2text-generation", "generated_from_trainer", "base_model:bigscience/mt0-base", "base_model:finetune:bigscience/mt0-base", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text2text-generation	2024-01-12T17:57:38Z	--- license: apache-2.0 base_model: bigscience/mt0-base tags: - generated_from_trainer model-index: - name: mt0-base_fine_lr5e-05_bs1_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # mt0-base_fine_lr5e-05_bs1_epoch5_wd0.01 This model is a fine-tuned version of [bigscience/mt0-base](https://huggingface.co/bigscience/mt0-base) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.0000 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 1 - eval_batch_size: 1 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:-----:\|:---------------:\| \| 0.006 \| 1.0 \| 8866 \| 0.0000 \| \| 0.0016 \| 2.0 \| 17732 \| 0.0000 \| \| 0.0003 \| 3.0 \| 26598 \| 0.0000 \| \| 0.007 \| 4.0 \| 35464 \| 0.0000 \| \| 0.0 \| 5.0 \| 44330 \| 0.0000 \| ### Framework versions - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
grimulkan/story-reverse-prompt-70b-rope8-32K-6bpw_h6_exl2	grimulkan	2024-01-12T19:18:42Z	4	0	transformers	[ "transformers", "llama", "text-generation", "license:llama2", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T18:12:52Z	--- license: llama2 --- This is a 6-bit quantization of [Story Reverse Prompt 70B 32K](https://huggingface.co/grimulkan/story-reverse-prompt-70b-rope8-32K-fp16). See that page for more details. This quantization fits in 48GB+24GB (36/24 split) or 3x24GB (16/17/20 split) using Exllamav2 @ 32k context.
LoneStriker/TenyxChat-7B-v1-6.0bpw-h6-exl2	LoneStriker	2024-01-12T19:03:20Z	6	0	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "tenyx-fine-tuning", "dpo", "tenyxchat", "conversational", "en", "arxiv:2305.18290", "arxiv:2309.11235", "arxiv:2306.05685", "arxiv:1803.05457", "arxiv:1905.07830", "arxiv:2009.03300", "arxiv:2109.07958", "arxiv:1907.10641", "arxiv:2110.14168", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T19:01:00Z	--- license: apache-2.0 language: - en library_name: transformers tags: - tenyx-fine-tuning - dpo - tenyxchat --- # TenyxChat: Language Model Alignment using Tenyx Fine-tuning Introducing TenyxChat, a series of ChatGPT-like models trained to function as useful assistants through preference tuning, using Tenyx's recently released advanced fine-tuning technology ([VentureBeat article](https://venturebeat.com/ai/tenyx-aims-to-fix-llms-catastrophic-forgetting-problem/)). Our first chat model in the series, TenyxChat-7B-v1, is trained using the [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290) framework on the open-source AI feedback dataset [UltraFeedback](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). We fine-tune [Openchat-3.5](https://arxiv.org/pdf/2309.11235.pdf) with our proprietary approach ([blog](https://www.tenyx.com/post/forgetting-and-toxicity-in-llms-a-deep-dive-on-fine-tuning-methods), [service](https://www.tenyx.com/fine-tuning)), which shows an increase in [MT-Bench](https://arxiv.org/abs/2306.05685), without a drop in performance of the model on other benchmarks. Our approach aims to mitigate forgetting in LLMs in a computationally efficient manner, thereby enabling continual fine-tuning capabilities without altering the pre-trained output distribution. TenyxChat-7B-v1 was trained using eight A100s (80GB) for two hours, with a training setup obtained from HuggingFaceH4 ([GitHub](https://github.com/huggingface/alignment-handbook)). # Model details - Model type: Fine-tuned 7B model for chat. - License: Apache 2.0 - Base model: OpenChat 3.5 ([https://huggingface.co/openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5)) - Demo: [spaces/tenyx/TenyxChat-7B-v1](https://huggingface.co/spaces/tenyx/TenyxChat-7B-v1) ## Usage Our model uses a simple chat template based on OpenChat 3.5. The chat template usage with a Hugging face generation example is shown below. ### Chat Template (Jinja) ```rust {{ bos_token }} {% for message in messages %} {% if message['role'] == 'user' %} {{ 'User:' + message['content'] + eos_token }} {% elif message['role'] == 'system' %} {{ 'System:' + message['content'] + eos_token }} {% elif message['role'] == 'assistant' %} {{ 'Assistant:' + message['content'] + eos_token }} {% endif %} {% if loop.last and add_generation_prompt %}\n{{ 'Assistant:' }}{% endif %}\n{% endfor %} ``` ### Hugging face Example ```python import torch from transformers import pipeline pipe = pipeline("text-generation", model="tenyx/TenyxChat-7B-v1", torch_dtype=torch.bfloat16, device_map="auto") messages = [ {"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate."}, {"role": "user", "content": "Hi. I would like to make a hotel booking."}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=512, do_sample=False) ``` ### Output ``` <s> System:You are a friendly chatbot who always responds in the style of a pirate.<\|end_of_turn\|> User:Hi. I would like to make a hotel booking.<\|end_of_turn\|> Assistant: Ahoy there me hearty! Arr, ye be lookin' fer a place to rest yer weary bones, eh? Well then, let's set sail on this grand adventure and find ye a swell place to stay! To begin, tell me the location ye be seekin' and the dates ye be lookin' to set sail. And don't ye worry, me matey, I'll be sure to find ye a place that'll make ye feel like a king or queen on land! ``` # Performance At the time of release (Jan 2024), TenyxChat-7B-v1 is the highest-ranked 7B chat model on the MT-Bench evaluation available for download and commercial use. We list here the benchmark results on several standard setups while comparing popular 7B models as baselines. ## MT-Bench MT-Bench is a benchmark made up of 80 high-quality multi-turn questions. These questions fall into eight categories: Writing, Roleplay, Reasoning, Math, Coding, Extraction, STEM, and Humanities. The chat models are rated using GPT-4 on a scale of 1 to 10, with higher values corresponding to better responses. \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| GPT-4* \| 8.95625 \| 9.02500 \| 8.990625 \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.75625 \| 8.103125 \| \| Starling-lm-7B-alpha \| 8.42500 \| 7.68750 \| 8.056250 \| \| OpenChat-3.5 \| 8.18125 \| 7.41250 \| 7.796875 \| \| GPT-3.5-turbo* \| 8.07500 \| 7.81250 \| 7.943750 \| \| OpenLLM Leader-7B** \| 8.05000 \| 7.61250 \| 7.831250 \| values reported on [lmsys](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) ChatBot Arena The [OpenLLM Leader](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) as of Jan 5, 2024 is the merge model available as [samir-fama/SamirGPT-v1](https://huggingface.co/samir-fama/SamirGPT-v1) ![hexplot.png](assets/hexplot.png) ### Comparison with additional Open LLM LeaderBoard models \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.756250 \| 8.103125 \| \| SamirGPT-v1 \| 8.05000 \| 7.612500 \| 7.831250 \| \| FernandoGPT-v1 \| 8.08125 \| 7.256250 \| 7.668750 \| \| Go-Bruins-v2 \| 8.13750 \| 7.150000 \| 7.643750 \| \| mistral_tv-neural-marconroni \| 7.76875 \| 6.987500 \| 7.378125 \| \| neuronovo-7B-v0.2 \| 7.73750 \| 6.662500 \| 7.200000 \| \| neural-chat-7b-v3-3 \| 7.39375 \| 5.881250 \| 6.637500 \| ## LM Evaluation - Open LLM Leaderboard We assess models on 7 benchmarks using the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). This setup is based of that used for [Open LLM Leaderboard.](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) - [AI2 Reasoning Challenge](https://arxiv.org/abs/1803.05457) (25-shot) - grade-school science questions. - [HellaSwag](https://arxiv.org/abs/1905.07830) (10-shot) - commonsense inference test. - [MMLU](https://arxiv.org/abs/2009.03300) (5-shot) - multitask accuracy test covering 57 tasks. - [TruthfulQA](https://arxiv.org/abs/2109.07958) (0-shot) - test measuring model's propensity to reproduce online falsehoods. - [Winogrande](https://arxiv.org/abs/1907.10641) (5-shot) - Winograd benchmark for commonsense reasoning. - [GSM8k](https://arxiv.org/abs/2110.14168) (5-shot) - grade school math word problems test. These benchmarks test reasoning and knowledge in various tasks in few-shot settings (higher scores are better). \| Model \| MMLU \| Winogrande \| GSM8k \| ARC \| HellaSwag \| TruthfulQA \| Average \| \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 63.6 \| 72.3 \| 69.0 \| 62.7 \| 66.6 \| 46.7 \| 63.48 \| \| Starling-7B-alpha \| 63.5 \| 72.1 \| 67.9 \| 61.1 \| 66.1 \| 42.1 \| 62.13 \| \| OpenChat-3.5 \| 63.6 \| 72.1 \| 68.2 \| 61.3 \| 65.2 \| 41.8 \| 62.03 \| \| Mistral-7B \| 62.4 \| 74.0 \| 38.1 \| 57.2 \| 62.8 \| 37.8 \| 55.38 \| \| OpenLLM Leader-7B \| 64.3 \| 78.7 \| 73.3 \| 66.6 \| 68.4 \| 58.5 \| 68.3 \| Note:* While the Open LLM Leaderboard indicates that these chat models perform less effectively compared to the leading 7B model, it's important to note that the leading model struggles in the multi-turn chat setting of MT-Bench (as demonstrated in our evaluation [above](#comparison-with-additional-open-llm-leaderboard-models)). In contrast, TenyxChat-7B-v1 demonstrates robustness against common fine-tuning challenges, such as catastrophic forgetting. This unique feature enables TenyxChat-7B-v1 to excel not only in chat benchmarks like MT-Bench, but also in a wider range of general reasoning benchmarks on the Open LLM Leaderboard. # Limitations TenyxChat-7B-v1, like other small-sized language models, has its own set of limitations. We haven’t fine-tuned the model explicitly to align with human safety preferences. Therefore, it is capable of producing undesirable outputs, particularly when adversarially prompted. From our observation, the model still tends to struggle with tasks that involve reasoning and math questions. In some instances, it might generate verbose or extraneous content. # License TenyxChat-7B-v1, similar to OpenChat 3.5, is distributed under the Apache License 2.0. # Citation If you use TenyxChat-7B for your research, cite us as ``` @misc{tenyxchat2024, title={TenyxChat: Language Model Alignment using Tenyx Fine-tuning}, author={Tenyx}, year={2024}, } ```
LoneStriker/TenyxChat-7B-v1-4.0bpw-h6-exl2	LoneStriker	2024-01-12T18:51:53Z	4	0	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "tenyx-fine-tuning", "dpo", "tenyxchat", "conversational", "en", "arxiv:2305.18290", "arxiv:2309.11235", "arxiv:2306.05685", "arxiv:1803.05457", "arxiv:1905.07830", "arxiv:2009.03300", "arxiv:2109.07958", "arxiv:1907.10641", "arxiv:2110.14168", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T18:50:14Z	--- license: apache-2.0 language: - en library_name: transformers tags: - tenyx-fine-tuning - dpo - tenyxchat --- # TenyxChat: Language Model Alignment using Tenyx Fine-tuning Introducing TenyxChat, a series of ChatGPT-like models trained to function as useful assistants through preference tuning, using Tenyx's recently released advanced fine-tuning technology ([VentureBeat article](https://venturebeat.com/ai/tenyx-aims-to-fix-llms-catastrophic-forgetting-problem/)). Our first chat model in the series, TenyxChat-7B-v1, is trained using the [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290) framework on the open-source AI feedback dataset [UltraFeedback](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). We fine-tune [Openchat-3.5](https://arxiv.org/pdf/2309.11235.pdf) with our proprietary approach ([blog](https://www.tenyx.com/post/forgetting-and-toxicity-in-llms-a-deep-dive-on-fine-tuning-methods), [service](https://www.tenyx.com/fine-tuning)), which shows an increase in [MT-Bench](https://arxiv.org/abs/2306.05685), without a drop in performance of the model on other benchmarks. Our approach aims to mitigate forgetting in LLMs in a computationally efficient manner, thereby enabling continual fine-tuning capabilities without altering the pre-trained output distribution. TenyxChat-7B-v1 was trained using eight A100s (80GB) for two hours, with a training setup obtained from HuggingFaceH4 ([GitHub](https://github.com/huggingface/alignment-handbook)). # Model details - Model type: Fine-tuned 7B model for chat. - License: Apache 2.0 - Base model: OpenChat 3.5 ([https://huggingface.co/openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5)) - Demo: [spaces/tenyx/TenyxChat-7B-v1](https://huggingface.co/spaces/tenyx/TenyxChat-7B-v1) ## Usage Our model uses a simple chat template based on OpenChat 3.5. The chat template usage with a Hugging face generation example is shown below. ### Chat Template (Jinja) ```rust {{ bos_token }} {% for message in messages %} {% if message['role'] == 'user' %} {{ 'User:' + message['content'] + eos_token }} {% elif message['role'] == 'system' %} {{ 'System:' + message['content'] + eos_token }} {% elif message['role'] == 'assistant' %} {{ 'Assistant:' + message['content'] + eos_token }} {% endif %} {% if loop.last and add_generation_prompt %}\n{{ 'Assistant:' }}{% endif %}\n{% endfor %} ``` ### Hugging face Example ```python import torch from transformers import pipeline pipe = pipeline("text-generation", model="tenyx/TenyxChat-7B-v1", torch_dtype=torch.bfloat16, device_map="auto") messages = [ {"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate."}, {"role": "user", "content": "Hi. I would like to make a hotel booking."}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=512, do_sample=False) ``` ### Output ``` <s> System:You are a friendly chatbot who always responds in the style of a pirate.<\|end_of_turn\|> User:Hi. I would like to make a hotel booking.<\|end_of_turn\|> Assistant: Ahoy there me hearty! Arr, ye be lookin' fer a place to rest yer weary bones, eh? Well then, let's set sail on this grand adventure and find ye a swell place to stay! To begin, tell me the location ye be seekin' and the dates ye be lookin' to set sail. And don't ye worry, me matey, I'll be sure to find ye a place that'll make ye feel like a king or queen on land! ``` # Performance At the time of release (Jan 2024), TenyxChat-7B-v1 is the highest-ranked 7B chat model on the MT-Bench evaluation available for download and commercial use. We list here the benchmark results on several standard setups while comparing popular 7B models as baselines. ## MT-Bench MT-Bench is a benchmark made up of 80 high-quality multi-turn questions. These questions fall into eight categories: Writing, Roleplay, Reasoning, Math, Coding, Extraction, STEM, and Humanities. The chat models are rated using GPT-4 on a scale of 1 to 10, with higher values corresponding to better responses. \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| GPT-4* \| 8.95625 \| 9.02500 \| 8.990625 \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.75625 \| 8.103125 \| \| Starling-lm-7B-alpha \| 8.42500 \| 7.68750 \| 8.056250 \| \| OpenChat-3.5 \| 8.18125 \| 7.41250 \| 7.796875 \| \| GPT-3.5-turbo* \| 8.07500 \| 7.81250 \| 7.943750 \| \| OpenLLM Leader-7B** \| 8.05000 \| 7.61250 \| 7.831250 \| values reported on [lmsys](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) ChatBot Arena The [OpenLLM Leader](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) as of Jan 5, 2024 is the merge model available as [samir-fama/SamirGPT-v1](https://huggingface.co/samir-fama/SamirGPT-v1) ![hexplot.png](assets/hexplot.png) ### Comparison with additional Open LLM LeaderBoard models \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.756250 \| 8.103125 \| \| SamirGPT-v1 \| 8.05000 \| 7.612500 \| 7.831250 \| \| FernandoGPT-v1 \| 8.08125 \| 7.256250 \| 7.668750 \| \| Go-Bruins-v2 \| 8.13750 \| 7.150000 \| 7.643750 \| \| mistral_tv-neural-marconroni \| 7.76875 \| 6.987500 \| 7.378125 \| \| neuronovo-7B-v0.2 \| 7.73750 \| 6.662500 \| 7.200000 \| \| neural-chat-7b-v3-3 \| 7.39375 \| 5.881250 \| 6.637500 \| ## LM Evaluation - Open LLM Leaderboard We assess models on 7 benchmarks using the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). This setup is based of that used for [Open LLM Leaderboard.](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) - [AI2 Reasoning Challenge](https://arxiv.org/abs/1803.05457) (25-shot) - grade-school science questions. - [HellaSwag](https://arxiv.org/abs/1905.07830) (10-shot) - commonsense inference test. - [MMLU](https://arxiv.org/abs/2009.03300) (5-shot) - multitask accuracy test covering 57 tasks. - [TruthfulQA](https://arxiv.org/abs/2109.07958) (0-shot) - test measuring model's propensity to reproduce online falsehoods. - [Winogrande](https://arxiv.org/abs/1907.10641) (5-shot) - Winograd benchmark for commonsense reasoning. - [GSM8k](https://arxiv.org/abs/2110.14168) (5-shot) - grade school math word problems test. These benchmarks test reasoning and knowledge in various tasks in few-shot settings (higher scores are better). \| Model \| MMLU \| Winogrande \| GSM8k \| ARC \| HellaSwag \| TruthfulQA \| Average \| \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 63.6 \| 72.3 \| 69.0 \| 62.7 \| 66.6 \| 46.7 \| 63.48 \| \| Starling-7B-alpha \| 63.5 \| 72.1 \| 67.9 \| 61.1 \| 66.1 \| 42.1 \| 62.13 \| \| OpenChat-3.5 \| 63.6 \| 72.1 \| 68.2 \| 61.3 \| 65.2 \| 41.8 \| 62.03 \| \| Mistral-7B \| 62.4 \| 74.0 \| 38.1 \| 57.2 \| 62.8 \| 37.8 \| 55.38 \| \| OpenLLM Leader-7B \| 64.3 \| 78.7 \| 73.3 \| 66.6 \| 68.4 \| 58.5 \| 68.3 \| Note:* While the Open LLM Leaderboard indicates that these chat models perform less effectively compared to the leading 7B model, it's important to note that the leading model struggles in the multi-turn chat setting of MT-Bench (as demonstrated in our evaluation [above](#comparison-with-additional-open-llm-leaderboard-models)). In contrast, TenyxChat-7B-v1 demonstrates robustness against common fine-tuning challenges, such as catastrophic forgetting. This unique feature enables TenyxChat-7B-v1 to excel not only in chat benchmarks like MT-Bench, but also in a wider range of general reasoning benchmarks on the Open LLM Leaderboard. # Limitations TenyxChat-7B-v1, like other small-sized language models, has its own set of limitations. We haven’t fine-tuned the model explicitly to align with human safety preferences. Therefore, it is capable of producing undesirable outputs, particularly when adversarially prompted. From our observation, the model still tends to struggle with tasks that involve reasoning and math questions. In some instances, it might generate verbose or extraneous content. # License TenyxChat-7B-v1, similar to OpenChat 3.5, is distributed under the Apache License 2.0. # Citation If you use TenyxChat-7B for your research, cite us as ``` @misc{tenyxchat2024, title={TenyxChat: Language Model Alignment using Tenyx Fine-tuning}, author={Tenyx}, year={2024}, } ```
virtsion/nilmformer_final_generic_prompt	virtsion	2024-01-12T18:48:31Z	0	0	peft	[ "peft", "safetensors", "arxiv:1910.09700", "base_model:meta-llama/Llama-2-7b-chat-hf", "base_model:adapter:meta-llama/Llama-2-7b-chat-hf", "region:us" ]	null	2024-01-12T18:48:29Z	--- library_name: peft base_model: meta-llama/Llama-2-7b-chat-hf --- # Model Card for Model ID <!-- Provide a quick summary of what the model is/does. --> ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> - Developed by: [More Information Needed] - Funded by [optional]: [More Information Needed] - Shared by [optional]: [More Information Needed] - Model type: [More Information Needed] - Language(s) (NLP): [More Information Needed] - License: [More Information Needed] - Finetuned from model [optional]: [More Information Needed] ### Model Sources [optional] <!-- Provide the basic links for the model. --> - Repository: [More Information Needed] - Paper [optional]: [More Information Needed] - Demo [optional]: [More Information Needed] ## Uses <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> [More Information Needed] ### Downstream Use [optional] <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app --> [More Information Needed] ### Out-of-Scope Use <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> [More Information Needed] ## Bias, Risks, and Limitations <!-- This section is meant to convey both technical and sociotechnical limitations. --> [More Information Needed] ### Recommendations <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. --> Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. ## How to Get Started with the Model Use the code below to get started with the model. [More Information Needed] ## Training Details ### Training Data <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> [More Information Needed] ### Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> #### Preprocessing [optional] [More Information Needed] #### Training Hyperparameters - Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision --> #### Speeds, Sizes, Times [optional] <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. --> [More Information Needed] ## Evaluation <!-- This section describes the evaluation protocols and provides the results. --> ### Testing Data, Factors & Metrics #### Testing Data <!-- This should link to a Dataset Card if possible. --> [More Information Needed] #### Factors <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. --> [More Information Needed] #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> [More Information Needed] ### Results [More Information Needed] #### Summary ## Model Examination [optional] <!-- Relevant interpretability work for the model goes here --> [More Information Needed] ## Environmental Impact <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly --> Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - Hardware Type: [More Information Needed] - Hours used: [More Information Needed] - Cloud Provider: [More Information Needed] - Compute Region: [More Information Needed] - Carbon Emitted: [More Information Needed] ## Technical Specifications [optional] ### Model Architecture and Objective [More Information Needed] ### Compute Infrastructure [More Information Needed] #### Hardware [More Information Needed] #### Software [More Information Needed] ## Citation [optional] <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> BibTeX: [More Information Needed] APA: [More Information Needed] ## Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> [More Information Needed] ## More Information [optional] [More Information Needed] ## Model Card Authors [optional] [More Information Needed] ## Model Card Contact [More Information Needed] ### Framework versions - PEFT 0.7.1
LoneStriker/TenyxChat-7B-v1-3.0bpw-h6-exl2	LoneStriker	2024-01-12T18:46:09Z	4	0	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "tenyx-fine-tuning", "dpo", "tenyxchat", "conversational", "en", "arxiv:2305.18290", "arxiv:2309.11235", "arxiv:2306.05685", "arxiv:1803.05457", "arxiv:1905.07830", "arxiv:2009.03300", "arxiv:2109.07958", "arxiv:1907.10641", "arxiv:2110.14168", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T18:44:50Z	--- license: apache-2.0 language: - en library_name: transformers tags: - tenyx-fine-tuning - dpo - tenyxchat --- # TenyxChat: Language Model Alignment using Tenyx Fine-tuning Introducing TenyxChat, a series of ChatGPT-like models trained to function as useful assistants through preference tuning, using Tenyx's recently released advanced fine-tuning technology ([VentureBeat article](https://venturebeat.com/ai/tenyx-aims-to-fix-llms-catastrophic-forgetting-problem/)). Our first chat model in the series, TenyxChat-7B-v1, is trained using the [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290) framework on the open-source AI feedback dataset [UltraFeedback](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). We fine-tune [Openchat-3.5](https://arxiv.org/pdf/2309.11235.pdf) with our proprietary approach ([blog](https://www.tenyx.com/post/forgetting-and-toxicity-in-llms-a-deep-dive-on-fine-tuning-methods), [service](https://www.tenyx.com/fine-tuning)), which shows an increase in [MT-Bench](https://arxiv.org/abs/2306.05685), without a drop in performance of the model on other benchmarks. Our approach aims to mitigate forgetting in LLMs in a computationally efficient manner, thereby enabling continual fine-tuning capabilities without altering the pre-trained output distribution. TenyxChat-7B-v1 was trained using eight A100s (80GB) for two hours, with a training setup obtained from HuggingFaceH4 ([GitHub](https://github.com/huggingface/alignment-handbook)). # Model details - Model type: Fine-tuned 7B model for chat. - License: Apache 2.0 - Base model: OpenChat 3.5 ([https://huggingface.co/openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5)) - Demo: [spaces/tenyx/TenyxChat-7B-v1](https://huggingface.co/spaces/tenyx/TenyxChat-7B-v1) ## Usage Our model uses a simple chat template based on OpenChat 3.5. The chat template usage with a Hugging face generation example is shown below. ### Chat Template (Jinja) ```rust {{ bos_token }} {% for message in messages %} {% if message['role'] == 'user' %} {{ 'User:' + message['content'] + eos_token }} {% elif message['role'] == 'system' %} {{ 'System:' + message['content'] + eos_token }} {% elif message['role'] == 'assistant' %} {{ 'Assistant:' + message['content'] + eos_token }} {% endif %} {% if loop.last and add_generation_prompt %}\n{{ 'Assistant:' }}{% endif %}\n{% endfor %} ``` ### Hugging face Example ```python import torch from transformers import pipeline pipe = pipeline("text-generation", model="tenyx/TenyxChat-7B-v1", torch_dtype=torch.bfloat16, device_map="auto") messages = [ {"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate."}, {"role": "user", "content": "Hi. I would like to make a hotel booking."}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=512, do_sample=False) ``` ### Output ``` <s> System:You are a friendly chatbot who always responds in the style of a pirate.<\|end_of_turn\|> User:Hi. I would like to make a hotel booking.<\|end_of_turn\|> Assistant: Ahoy there me hearty! Arr, ye be lookin' fer a place to rest yer weary bones, eh? Well then, let's set sail on this grand adventure and find ye a swell place to stay! To begin, tell me the location ye be seekin' and the dates ye be lookin' to set sail. And don't ye worry, me matey, I'll be sure to find ye a place that'll make ye feel like a king or queen on land! ``` # Performance At the time of release (Jan 2024), TenyxChat-7B-v1 is the highest-ranked 7B chat model on the MT-Bench evaluation available for download and commercial use. We list here the benchmark results on several standard setups while comparing popular 7B models as baselines. ## MT-Bench MT-Bench is a benchmark made up of 80 high-quality multi-turn questions. These questions fall into eight categories: Writing, Roleplay, Reasoning, Math, Coding, Extraction, STEM, and Humanities. The chat models are rated using GPT-4 on a scale of 1 to 10, with higher values corresponding to better responses. \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| GPT-4* \| 8.95625 \| 9.02500 \| 8.990625 \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.75625 \| 8.103125 \| \| Starling-lm-7B-alpha \| 8.42500 \| 7.68750 \| 8.056250 \| \| OpenChat-3.5 \| 8.18125 \| 7.41250 \| 7.796875 \| \| GPT-3.5-turbo* \| 8.07500 \| 7.81250 \| 7.943750 \| \| OpenLLM Leader-7B** \| 8.05000 \| 7.61250 \| 7.831250 \| values reported on [lmsys](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) ChatBot Arena The [OpenLLM Leader](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) as of Jan 5, 2024 is the merge model available as [samir-fama/SamirGPT-v1](https://huggingface.co/samir-fama/SamirGPT-v1) ![hexplot.png](assets/hexplot.png) ### Comparison with additional Open LLM LeaderBoard models \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.756250 \| 8.103125 \| \| SamirGPT-v1 \| 8.05000 \| 7.612500 \| 7.831250 \| \| FernandoGPT-v1 \| 8.08125 \| 7.256250 \| 7.668750 \| \| Go-Bruins-v2 \| 8.13750 \| 7.150000 \| 7.643750 \| \| mistral_tv-neural-marconroni \| 7.76875 \| 6.987500 \| 7.378125 \| \| neuronovo-7B-v0.2 \| 7.73750 \| 6.662500 \| 7.200000 \| \| neural-chat-7b-v3-3 \| 7.39375 \| 5.881250 \| 6.637500 \| ## LM Evaluation - Open LLM Leaderboard We assess models on 7 benchmarks using the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). This setup is based of that used for [Open LLM Leaderboard.](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) - [AI2 Reasoning Challenge](https://arxiv.org/abs/1803.05457) (25-shot) - grade-school science questions. - [HellaSwag](https://arxiv.org/abs/1905.07830) (10-shot) - commonsense inference test. - [MMLU](https://arxiv.org/abs/2009.03300) (5-shot) - multitask accuracy test covering 57 tasks. - [TruthfulQA](https://arxiv.org/abs/2109.07958) (0-shot) - test measuring model's propensity to reproduce online falsehoods. - [Winogrande](https://arxiv.org/abs/1907.10641) (5-shot) - Winograd benchmark for commonsense reasoning. - [GSM8k](https://arxiv.org/abs/2110.14168) (5-shot) - grade school math word problems test. These benchmarks test reasoning and knowledge in various tasks in few-shot settings (higher scores are better). \| Model \| MMLU \| Winogrande \| GSM8k \| ARC \| HellaSwag \| TruthfulQA \| Average \| \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 63.6 \| 72.3 \| 69.0 \| 62.7 \| 66.6 \| 46.7 \| 63.48 \| \| Starling-7B-alpha \| 63.5 \| 72.1 \| 67.9 \| 61.1 \| 66.1 \| 42.1 \| 62.13 \| \| OpenChat-3.5 \| 63.6 \| 72.1 \| 68.2 \| 61.3 \| 65.2 \| 41.8 \| 62.03 \| \| Mistral-7B \| 62.4 \| 74.0 \| 38.1 \| 57.2 \| 62.8 \| 37.8 \| 55.38 \| \| OpenLLM Leader-7B \| 64.3 \| 78.7 \| 73.3 \| 66.6 \| 68.4 \| 58.5 \| 68.3 \| Note:* While the Open LLM Leaderboard indicates that these chat models perform less effectively compared to the leading 7B model, it's important to note that the leading model struggles in the multi-turn chat setting of MT-Bench (as demonstrated in our evaluation [above](#comparison-with-additional-open-llm-leaderboard-models)). In contrast, TenyxChat-7B-v1 demonstrates robustness against common fine-tuning challenges, such as catastrophic forgetting. This unique feature enables TenyxChat-7B-v1 to excel not only in chat benchmarks like MT-Bench, but also in a wider range of general reasoning benchmarks on the Open LLM Leaderboard. # Limitations TenyxChat-7B-v1, like other small-sized language models, has its own set of limitations. We haven’t fine-tuned the model explicitly to align with human safety preferences. Therefore, it is capable of producing undesirable outputs, particularly when adversarially prompted. From our observation, the model still tends to struggle with tasks that involve reasoning and math questions. In some instances, it might generate verbose or extraneous content. # License TenyxChat-7B-v1, similar to OpenChat 3.5, is distributed under the Apache License 2.0. # Citation If you use TenyxChat-7B for your research, cite us as ``` @misc{tenyxchat2024, title={TenyxChat: Language Model Alignment using Tenyx Fine-tuning}, author={Tenyx}, year={2024}, } ```
jysssacc/opt-350m_lora_lr0.0005_bs10_epoch5_wd0.01	jysssacc	2024-01-12T18:45:56Z	1	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:facebook/opt-350m", "base_model:adapter:facebook/opt-350m", "license:other", "region:us" ]	null	2024-01-12T18:45:19Z	--- license: other library_name: peft tags: - generated_from_trainer base_model: facebook/opt-350m model-index: - name: opt-350m_lora_lr0.0005_bs10_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # opt-350m_lora_lr0.0005_bs10_epoch5_wd0.01 This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.3942 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0005 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| No log \| 1.0 \| 63 \| 3.4105 \| \| 3.7109 \| 2.0 \| 126 \| 3.4126 \| \| 3.7109 \| 3.0 \| 189 \| 3.4107 \| \| 3.4287 \| 4.0 \| 252 \| 3.3937 \| \| 3.3121 \| 5.0 \| 315 \| 3.3942 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
TheBloke/UNA-TheBeagle-7B-v1-GPTQ	TheBloke	2024-01-12T18:42:59Z	128	3	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "generated_from_trainer", "dataset:jondurbin/bagel-v0.3", "base_model:fblgit/UNA-TheBeagle-7b-v1", "base_model:quantized:fblgit/UNA-TheBeagle-7b-v1", "license:cc-by-nc-nd-4.0", "autotrain_compatible", "text-generation-inference", "4-bit", "gptq", "region:us" ]	text-generation	2024-01-12T18:19:11Z	--- base_model: fblgit/UNA-TheBeagle-7b-v1 datasets: - jondurbin/bagel-v0.3 inference: false library_name: transformers license: cc-by-nc-nd-4.0 model-index: - name: UNA-TheBeagle-7b-v1 results: [] model_creator: FBL model_name: UNA TheBeagle 7B v1 model_type: mistral prompt_template: '{prompt} ' quantized_by: TheBloke tags: - generated_from_trainer --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # UNA TheBeagle 7B v1 - GPTQ - Model creator: [FBL](https://huggingface.co/fblgit) - Original model: [UNA TheBeagle 7B v1](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1) <!-- description start --> # Description This repo contains GPTQ model files for [FBL's UNA TheBeagle 7B v1](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1). Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them. <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF) * [FBL's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: Unknown ``` {prompt} ``` <!-- prompt-template end --> <!-- README_GPTQ.md-compatible clients start --> ## Known compatible clients / servers GPTQ models are currently supported on Linux (NVidia/AMD) and Windows (NVidia only). macOS users: please use GGUF models. These GPTQ models are known to work in the following inference servers/webuis. - [text-generation-webui](https://github.com/oobabooga/text-generation-webui) - [KoboldAI United](https://github.com/henk717/koboldai) - [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui) - [Hugging Face Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) This may not be a complete list; if you know of others, please let me know! <!-- README_GPTQ.md-compatible clients end --> <!-- README_GPTQ.md-provided-files start --> ## Provided files, and GPTQ parameters Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements. Each separate quant is in a different branch. See below for instructions on fetching from different branches. Most GPTQ files are made with AutoGPTQ. Mistral models are currently made with Transformers. <details> <summary>Explanation of GPTQ parameters</summary> - Bits: The bit size of the quantised model. - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value. - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now. - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy. - GPTQ dataset: The calibration dataset used during quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ calibration dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s). - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences. - ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama and Mistral models in 4-bit. </details> \| Branch \| Bits \| GS \| Act Order \| Damp % \| GPTQ Dataset \| Seq Len \| Size \| ExLlama \| Desc \| \| ------ \| ---- \| -- \| --------- \| ------ \| ------------ \| ------- \| ---- \| ------- \| ---- \| \| [main](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ/tree/main) \| 4 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.16 GB \| Yes \| 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. \| \| [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ/tree/gptq-4bit-32g-actorder_True) \| 4 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.57 GB \| Yes \| 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. \| \| [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ/tree/gptq-8bit--1g-actorder_True) \| 8 \| None \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 7.52 GB \| No \| 8-bit, with Act Order. No group size, to lower VRAM requirements. \| \| [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ/tree/gptq-8bit-128g-actorder_True) \| 8 \| 128 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 7.68 GB \| No \| 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. \| \| [gptq-8bit-32g-actorder_True](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ/tree/gptq-8bit-32g-actorder_True) \| 8 \| 32 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 8.17 GB \| No \| 8-bit, with group size 32g and Act Order for maximum inference quality. \| \| [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ/tree/gptq-4bit-64g-actorder_True) \| 4 \| 64 \| Yes \| 0.1 \| [VMware Open Instruct](https://huggingface.co/datasets/VMware/open-instruct/viewer/) \| 4096 \| 4.29 GB \| Yes \| 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. \| <!-- README_GPTQ.md-provided-files end --> <!-- README_GPTQ.md-download-from-branches start --> ## How to download, including from branches ### In text-generation-webui To download from the `main` branch, enter `TheBloke/UNA-TheBeagle-7B-v1-GPTQ` in the "Download model" box. To download from another branch, add `:branchname` to the end of the download name, eg `TheBloke/UNA-TheBeagle-7B-v1-GPTQ:gptq-4bit-32g-actorder_True` ### From the command line I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` To download the `main` branch to a folder called `UNA-TheBeagle-7B-v1-GPTQ`: ```shell mkdir UNA-TheBeagle-7B-v1-GPTQ huggingface-cli download TheBloke/UNA-TheBeagle-7B-v1-GPTQ --local-dir UNA-TheBeagle-7B-v1-GPTQ --local-dir-use-symlinks False ``` To download from a different branch, add the `--revision` parameter: ```shell mkdir UNA-TheBeagle-7B-v1-GPTQ huggingface-cli download TheBloke/UNA-TheBeagle-7B-v1-GPTQ --revision gptq-4bit-32g-actorder_True --local-dir UNA-TheBeagle-7B-v1-GPTQ --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage</summary> If you remove the `--local-dir-use-symlinks False` parameter, the files will instead be stored in the central Hugging Face cache directory (default location on Linux is: `~/.cache/huggingface`), and symlinks will be added to the specified `--local-dir`, pointing to their real location in the cache. This allows for interrupted downloads to be resumed, and allows you to quickly clone the repo to multiple places on disk without triggering a download again. The downside, and the reason why I don't list that as the default option, is that the files are then hidden away in a cache folder and it's harder to know where your disk space is being used, and to clear it up if/when you want to remove a download model. The cache location can be changed with the `HF_HOME` environment variable, and/or the `--cache-dir` parameter to `huggingface-cli`. For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell mkdir UNA-TheBeagle-7B-v1-GPTQ HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/UNA-TheBeagle-7B-v1-GPTQ --local-dir UNA-TheBeagle-7B-v1-GPTQ --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> ### With `git` (not recommended) To clone a specific branch with `git`, use a command like this: ```shell git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ ``` Note that using Git with HF repos is strongly discouraged. It will be much slower than using `huggingface-hub`, and will use twice as much disk space as it has to store the model files twice (it stores every byte both in the intended target folder, and again in the `.git` folder as a blob.) <!-- README_GPTQ.md-download-from-branches end --> <!-- README_GPTQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui) Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the Model tab. 2. Under Download custom model or LoRA, enter `TheBloke/UNA-TheBeagle-7B-v1-GPTQ`. - To download from a specific branch, enter for example `TheBloke/UNA-TheBeagle-7B-v1-GPTQ:gptq-4bit-32g-actorder_True` - see Provided Files above for the list of branches for each option. 3. Click Download. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to Model. 6. In the Model dropdown, choose the model you just downloaded: `UNA-TheBeagle-7B-v1-GPTQ` 7. The model will automatically load, and is now ready for use! 8. If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right. - Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file `quantize_config.json`. 9. Once you're ready, click the Text Generation tab and enter a prompt to get started! <!-- README_GPTQ.md-text-generation-webui end --> <!-- README_GPTQ.md-use-from-tgi start --> ## Serving this model from Text Generation Inference (TGI) It's recommended to use TGI version 1.1.0 or later. The official Docker container is: `ghcr.io/huggingface/text-generation-inference:1.1.0` Example Docker parameters: ```shell --model-id TheBloke/UNA-TheBeagle-7B-v1-GPTQ --port 3000 --quantize gptq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 ``` Example Python code for interfacing with TGI (requires huggingface-hub 0.17.0 or later): ```shell pip3 install huggingface-hub ``` ```python from huggingface_hub import InferenceClient endpoint_url = "https://your-endpoint-url-here" prompt = "Tell me about AI" prompt_template=f'''{prompt} ''' client = InferenceClient(endpoint_url) response = client.text_generation( prompt_template, max_new_tokens=128, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(f"Model output: {response}") ``` <!-- README_GPTQ.md-use-from-tgi end --> <!-- README_GPTQ.md-use-from-python start --> ## Python code example: inference from this GPTQ model ### Install the necessary packages Requires: Transformers 4.33.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later. ```shell pip3 install --upgrade transformers optimum # If using PyTorch 2.1 + CUDA 12.x: pip3 install --upgrade auto-gptq # or, if using PyTorch 2.1 + CUDA 11.x: pip3 install --upgrade auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ ``` If you are using PyTorch 2.0, you will need to install AutoGPTQ from source. Likewise if you have problems with the pre-built wheels, you should try building from source: ```shell pip3 uninstall -y auto-gptq git clone https://github.com/PanQiWei/AutoGPTQ cd AutoGPTQ git checkout v0.5.1 pip3 install . ``` ### Example Python code ```python from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline model_name_or_path = "TheBloke/UNA-TheBeagle-7B-v1-GPTQ" # To use a different branch, change revision # For example: revision="gptq-4bit-32g-actorder_True" model = AutoModelForCausalLM.from_pretrained(model_name_or_path, device_map="auto", trust_remote_code=False, revision="main") tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True) prompt = "Write a story about llamas" system_message = "You are a story writing assistant" prompt_template=f'''{prompt} ''' print("\n\n* Generate:") input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda() output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512) print(tokenizer.decode(output[0])) # Inference can also be done using transformers' pipeline print("* Pipeline:") pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(pipe(prompt_template)[0]['generated_text']) ``` <!-- README_GPTQ.md-use-from-python end --> <!-- README_GPTQ.md-compatibility start --> ## Compatibility The files provided are tested to work with Transformers. For non-Mistral models, AutoGPTQ can also be used directly. [ExLlama](https://github.com/turboderp/exllama) is compatible with Llama architecture models (including Mistral, Yi, DeepSeek, SOLAR, etc) in 4-bit. Please see the Provided Files table above for per-file compatibility. For a list of clients/servers, please see "Known compatible clients / servers", above. <!-- README_GPTQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: FBL's UNA TheBeagle 7B v1 -- In the Love Memory of my "LoLa" -- # UNA-TheBeagle-7b-v1 TheBeagle, a model of 7B parameters trained on The Bagel dataset. DPO & UNA applied over a set of curated DPO Pairs. - Scored #1 on the HF Leaderboard, dramatic scores!!! 73 ARC, and very well balanced! The dataset was generated using the original bagel code, including the decontamination step. As base model, we used the latest Intel's neural-chat model. It performs very good in many tasks, but its always better that you play with it by yourself. ![TheBeagle](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1/resolve/main/TheBeagle.png) ## Evaluations Ran with VLLM so expect them to dont be exactly as the one's shown in the board, but not too far :) ``` vllm (pretrained=fblgit/UNA-TheBeagle-7b-v1,dtype=auto,tensor_parallel_size=1,gpu_memory_utilization=0.8,data_parallel_size=8,trust_remote_code=True), gen_kwargs: (None), limit: None, num_fewshot: None, batch_size: 32 \| Tasks \|Version\| Filter \|n-shot\| Metric \|Value \| \|Stderr\| \|--------------\|-------\|----------\|-----:\|-----------\|-----:\|---\|-----:\| \|arc_challenge \|Yaml \|none \| 25\|acc \|0.7090\|± \|0.0133\| \| \| \|none \| 25\|acc_norm \|0.7329\|± \|0.0129\| \|gsm8k \|Yaml \|get-answer\| 5\|exact_match\|0.7210\|± \|0.0124\| \|hellaswag \|Yaml \|none \| 10\|acc \|0.7202\|± \|0.0045\| \| \| \|none \| 10\|acc_norm \|0.8792\|± \|0.0033\| \|truthfulqa_mc2\|Yaml \|none \| 0\|acc \|0.7062\|± \|0.0151\| \|winogrande \|Yaml \|none \| 5\|acc \|0.8366\|± \|0.0104\| ``` ## UNA Details For this release, we only applied UNA thru the perceptrons. It was done at a 3.5e-7 speed, and the training loop code is also the original one of the bagel and transformers-4.35.2-UNA ## Prompt Im not entirely sure of it, as we used the vanilla version of the bagel training code. But a good model should be able to generalize with different prompt formats, so feel free to give it a shot. ## Citations Remember if you use UNA's models, cite it in your model card. ## Limitations Not for commercial use, and only for academic & research purposes.
TheBloke/UNA-TheBeagle-7B-v1-GGUF	TheBloke	2024-01-12T18:27:59Z	610	18	transformers	[ "transformers", "gguf", "mistral", "generated_from_trainer", "dataset:jondurbin/bagel-v0.3", "base_model:fblgit/UNA-TheBeagle-7b-v1", "base_model:quantized:fblgit/UNA-TheBeagle-7b-v1", "license:cc-by-nc-nd-4.0", "region:us" ]	null	2024-01-12T18:19:11Z	--- base_model: fblgit/UNA-TheBeagle-7b-v1 datasets: - jondurbin/bagel-v0.3 inference: false library_name: transformers license: cc-by-nc-nd-4.0 model-index: - name: UNA-TheBeagle-7b-v1 results: [] model_creator: FBL model_name: UNA TheBeagle 7B v1 model_type: mistral prompt_template: '{prompt} ' quantized_by: TheBloke tags: - generated_from_trainer --- <!-- markdownlint-disable MD041 --> <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # UNA TheBeagle 7B v1 - GGUF - Model creator: [FBL](https://huggingface.co/fblgit) - Original model: [UNA TheBeagle 7B v1](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1) <!-- description start --> ## Description This repo contains GGUF format model files for [FBL's UNA TheBeagle 7B v1](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1). <!-- description end --> <!-- README_GGUF.md-about-gguf start --> ### About GGUF GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp. Here is an incomplete list of clients and libraries that are known to support GGUF: * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option. * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration. * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling. * [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel. * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023. * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection. * [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration. * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server. * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use. * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models. <!-- README_GGUF.md-about-gguf end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF) * [FBL's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: Unknown ``` {prompt} ``` <!-- prompt-template end --> <!-- compatibility_gguf start --> ## Compatibility These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) They are also compatible with many third party UIs and libraries - please see the list at the top of this README. ## Explanation of quantisation methods <details> <summary>Click to see details</summary> The new methods available are: * GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw) * GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw. * GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw. * GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw * GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw Refer to the Provided Files table below to see what files use which methods, and how. </details> <!-- compatibility_gguf end --> <!-- README_GGUF.md-provided-files start --> ## Provided files \| Name \| Quant method \| Bits \| Size \| Max RAM required \| Use case \| \| ---- \| ---- \| ---- \| ---- \| ---- \| ----- \| \| [una-thebeagle-7b-v1.Q2_K.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q2_K.gguf) \| Q2_K \| 2 \| 2.70 GB\| 5.20 GB \| smallest, significant quality loss - not recommended for most purposes \| \| [una-thebeagle-7b-v1.Q3_K_S.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q3_K_S.gguf) \| Q3_K_S \| 3 \| 3.16 GB\| 5.66 GB \| very small, high quality loss \| \| [una-thebeagle-7b-v1.Q3_K_M.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q3_K_M.gguf) \| Q3_K_M \| 3 \| 3.52 GB\| 6.02 GB \| very small, high quality loss \| \| [una-thebeagle-7b-v1.Q3_K_L.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q3_K_L.gguf) \| Q3_K_L \| 3 \| 3.82 GB\| 6.32 GB \| small, substantial quality loss \| \| [una-thebeagle-7b-v1.Q4_0.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q4_0.gguf) \| Q4_0 \| 4 \| 4.11 GB\| 6.61 GB \| legacy; small, very high quality loss - prefer using Q3_K_M \| \| [una-thebeagle-7b-v1.Q4_K_S.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q4_K_S.gguf) \| Q4_K_S \| 4 \| 4.14 GB\| 6.64 GB \| small, greater quality loss \| \| [una-thebeagle-7b-v1.Q4_K_M.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q4_K_M.gguf) \| Q4_K_M \| 4 \| 4.37 GB\| 6.87 GB \| medium, balanced quality - recommended \| \| [una-thebeagle-7b-v1.Q5_0.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q5_0.gguf) \| Q5_0 \| 5 \| 5.00 GB\| 7.50 GB \| legacy; medium, balanced quality - prefer using Q4_K_M \| \| [una-thebeagle-7b-v1.Q5_K_S.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q5_K_S.gguf) \| Q5_K_S \| 5 \| 5.00 GB\| 7.50 GB \| large, low quality loss - recommended \| \| [una-thebeagle-7b-v1.Q5_K_M.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q5_K_M.gguf) \| Q5_K_M \| 5 \| 5.13 GB\| 7.63 GB \| large, very low quality loss - recommended \| \| [una-thebeagle-7b-v1.Q6_K.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q6_K.gguf) \| Q6_K \| 6 \| 5.94 GB\| 8.44 GB \| very large, extremely low quality loss \| \| [una-thebeagle-7b-v1.Q8_0.gguf](https://huggingface.co/TheBloke/UNA-TheBeagle-7B-v1-GGUF/blob/main/una-thebeagle-7b-v1.Q8_0.gguf) \| Q8_0 \| 8 \| 7.70 GB\| 10.20 GB \| very large, extremely low quality loss - not recommended \| Note: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead. <!-- README_GGUF.md-provided-files end --> <!-- README_GGUF.md-how-to-download start --> ## How to download GGUF files Note for manual downloaders: You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file. The following clients/libraries will automatically download models for you, providing a list of available models to choose from: * LM Studio * LoLLMS Web UI * Faraday.dev ### In `text-generation-webui` Under Download Model, you can enter the model repo: TheBloke/UNA-TheBeagle-7B-v1-GGUF and below it, a specific filename to download, such as: una-thebeagle-7b-v1.Q4_K_M.gguf. Then click Download. ### On the command line, including multiple files at once I recommend using the `huggingface-hub` Python library: ```shell pip3 install huggingface-hub ``` Then you can download any individual model file to the current directory, at high speed, with a command like this: ```shell huggingface-cli download TheBloke/UNA-TheBeagle-7B-v1-GGUF una-thebeagle-7b-v1.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` <details> <summary>More advanced huggingface-cli download usage (click to read)</summary> You can also download multiple files at once with a pattern: ```shell huggingface-cli download TheBloke/UNA-TheBeagle-7B-v1-GGUF --local-dir . --local-dir-use-symlinks False --include='Q4_Kgguf' ``` For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli). To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`: ```shell pip3 install hf_transfer ``` And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`: ```shell HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/UNA-TheBeagle-7B-v1-GGUF una-thebeagle-7b-v1.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False ``` Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command. </details> <!-- README_GGUF.md-how-to-download end --> <!-- README_GGUF.md-how-to-run start --> ## Example `llama.cpp` command Make sure you are using `llama.cpp` from commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later. ```shell ./main -ngl 35 -m una-thebeagle-7b-v1.Q4_K_M.gguf --color -c 32768 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "{prompt}" ``` Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration. Change `-c 32768` to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically. Note that longer sequence lengths require much more resources, so you may need to reduce this value. If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins` For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md) ## How to run in `text-generation-webui` Further instructions can be found in the text-generation-webui documentation, here: [text-generation-webui/docs/04 ‐ Model Tab.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/04%20%E2%80%90%20Model%20Tab.md#llamacpp). ## How to run from Python code You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) or [ctransformers](https://github.com/marella/ctransformers) libraries. Note that at the time of writing (Nov 27th 2023), ctransformers has not been updated for some time and is not compatible with some recent models. Therefore I recommend you use llama-cpp-python. ### How to load this model in Python code, using llama-cpp-python For full documentation, please see: [llama-cpp-python docs](https://abetlen.github.io/llama-cpp-python/). #### First install the package Run one of the following commands, according to your system: ```shell # Base ctransformers with no GPU acceleration pip install llama-cpp-python # With NVidia CUDA acceleration CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install llama-cpp-python # Or with OpenBLAS acceleration CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python # Or with CLBLast acceleration CMAKE_ARGS="-DLLAMA_CLBLAST=on" pip install llama-cpp-python # Or with AMD ROCm GPU acceleration (Linux only) CMAKE_ARGS="-DLLAMA_HIPBLAS=on" pip install llama-cpp-python # Or with Metal GPU acceleration for macOS systems only CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python # In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA: $env:CMAKE_ARGS = "-DLLAMA_OPENBLAS=on" pip install llama-cpp-python ``` #### Simple llama-cpp-python example code ```python from llama_cpp import Llama # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system. llm = Llama( model_path="./una-thebeagle-7b-v1.Q4_K_M.gguf", # Download the model file first n_ctx=32768, # The max sequence length to use - note that longer sequence lengths require much more resources n_threads=8, # The number of CPU threads to use, tailor to your system and the resulting performance n_gpu_layers=35 # The number of layers to offload to GPU, if you have GPU acceleration available ) # Simple inference example output = llm( "{prompt}", # Prompt max_tokens=512, # Generate up to 512 tokens stop=["</s>"], # Example stop token - not necessarily correct for this specific model! Please check before using. echo=True # Whether to echo the prompt ) # Chat Completion API llm = Llama(model_path="./una-thebeagle-7b-v1.Q4_K_M.gguf", chat_format="llama-2") # Set chat_format according to the model you are using llm.create_chat_completion( messages = [ {"role": "system", "content": "You are a story writing assistant."}, { "role": "user", "content": "Write a story about llamas." } ] ) ``` ## How to use with LangChain Here are guides on using llama-cpp-python and ctransformers with LangChain: * [LangChain + llama-cpp-python](https://python.langchain.com/docs/integrations/llms/llamacpp) * [LangChain + ctransformers](https://python.langchain.com/docs/integrations/providers/ctransformers) <!-- README_GGUF.md-how-to-run end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI Special thanks to: Aemon Algiz. Patreon special mentions: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> <!-- original-model-card start --> # Original model card: FBL's UNA TheBeagle 7B v1 -- In the Love Memory of my "LoLa" -- # UNA-TheBeagle-7b-v1 TheBeagle, a model of 7B parameters trained on The Bagel dataset. DPO & UNA applied over a set of curated DPO Pairs. - Scored #1 on the HF Leaderboard, dramatic scores!!! 73 ARC, and very well balanced! The dataset was generated using the original bagel code, including the decontamination step. As base model, we used the latest Intel's neural-chat model. It performs very good in many tasks, but its always better that you play with it by yourself. ![TheBeagle](https://huggingface.co/fblgit/UNA-TheBeagle-7b-v1/resolve/main/TheBeagle.png) ## Evaluations Ran with VLLM so expect them to dont be exactly as the one's shown in the board, but not too far :) ``` vllm (pretrained=fblgit/UNA-TheBeagle-7b-v1,dtype=auto,tensor_parallel_size=1,gpu_memory_utilization=0.8,data_parallel_size=8,trust_remote_code=True), gen_kwargs: (None), limit: None, num_fewshot: None, batch_size: 32 \| Tasks \|Version\| Filter \|n-shot\| Metric \|Value \| \|Stderr\| \|--------------\|-------\|----------\|-----:\|-----------\|-----:\|---\|-----:\| \|arc_challenge \|Yaml \|none \| 25\|acc \|0.7090\|± \|0.0133\| \| \| \|none \| 25\|acc_norm \|0.7329\|± \|0.0129\| \|gsm8k \|Yaml \|get-answer\| 5\|exact_match\|0.7210\|± \|0.0124\| \|hellaswag \|Yaml \|none \| 10\|acc \|0.7202\|± \|0.0045\| \| \| \|none \| 10\|acc_norm \|0.8792\|± \|0.0033\| \|truthfulqa_mc2\|Yaml \|none \| 0\|acc \|0.7062\|± \|0.0151\| \|winogrande \|Yaml \|none \| 5\|acc \|0.8366\|± \|0.0104\| ``` ## UNA Details For this release, we only applied UNA thru the perceptrons. It was done at a 3.5e-7 speed, and the training loop code is also the original one of the bagel and transformers-4.35.2-UNA ## Prompt Im not entirely sure of it, as we used the vanilla version of the bagel training code. But a good model should be able to generalize with different prompt formats, so feel free to give it a shot. ## Citations Remember if you use UNA's models, cite it in your model card. ## Limitations Not for commercial use, and only for academic & research purposes. <!-- original-model-card end -->
Perselope/Taxi-v3_v5	Perselope	2024-01-12T18:19:32Z	0	0	null	[ "Taxi-v3", "q-learning", "reinforcement-learning", "custom-implementation", "model-index", "region:us" ]	reinforcement-learning	2024-01-12T18:19:27Z	--- tags: - Taxi-v3 - q-learning - reinforcement-learning - custom-implementation model-index: - name: Taxi-v3_v5 results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: Taxi-v3 type: Taxi-v3 metrics: - type: mean_reward value: 7.96 +/- 2.30 name: mean_reward verified: false --- # Q-Learning Agent playing1 Taxi-v3 This is a trained model of a Q-Learning agent playing Taxi-v3 . ## Usage model = load_from_hub(repo_id="Perselope/Taxi-v3_v5", filename="q-learning.pkl") # Don't forget to check if you need to add additional attributes (is_slippery=False etc) env = gym.make(model["env_id"])
tenyx/TenyxChat-7B-v1	tenyx	2024-01-12T18:16:23Z	1,400	25	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "tenyx-fine-tuning", "dpo", "tenyxchat", "conversational", "en", "arxiv:2305.18290", "arxiv:2309.11235", "arxiv:2306.05685", "arxiv:1803.05457", "arxiv:1905.07830", "arxiv:2009.03300", "arxiv:2109.07958", "arxiv:1907.10641", "arxiv:2110.14168", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-05T23:18:34Z	--- license: apache-2.0 language: - en library_name: transformers tags: - tenyx-fine-tuning - dpo - tenyxchat --- # TenyxChat: Language Model Alignment using Tenyx Fine-tuning Introducing TenyxChat, a series of ChatGPT-like models trained to function as useful assistants through preference tuning, using Tenyx's recently released advanced fine-tuning technology ([VentureBeat article](https://venturebeat.com/ai/tenyx-aims-to-fix-llms-catastrophic-forgetting-problem/)). Our first chat model in the series, TenyxChat-7B-v1, is trained using the [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290) framework on the open-source AI feedback dataset [UltraFeedback](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized). We fine-tune [Openchat-3.5](https://arxiv.org/pdf/2309.11235.pdf) with our proprietary approach ([blog](https://www.tenyx.com/post/forgetting-and-toxicity-in-llms-a-deep-dive-on-fine-tuning-methods), [service](https://www.tenyx.com/fine-tuning)), which shows an increase in [MT-Bench](https://arxiv.org/abs/2306.05685), without a drop in performance of the model on other benchmarks. Our approach aims to mitigate forgetting in LLMs in a computationally efficient manner, thereby enabling continual fine-tuning capabilities without altering the pre-trained output distribution. TenyxChat-7B-v1 was trained using eight A100s (80GB) for two hours, with a training setup obtained from HuggingFaceH4 ([GitHub](https://github.com/huggingface/alignment-handbook)). # Model details - Model type: Fine-tuned 7B model for chat. - License: Apache 2.0 - Base model: OpenChat 3.5 ([https://huggingface.co/openchat/openchat_3.5](https://huggingface.co/openchat/openchat_3.5)) - Demo: [spaces/tenyx/TenyxChat-7B-v1](https://huggingface.co/spaces/tenyx/TenyxChat-7B-v1) ## Usage Our model uses a simple chat template based on OpenChat 3.5. The chat template usage with a Hugging face generation example is shown below. ### Chat Template (Jinja) ```rust {{ bos_token }} {% for message in messages %} {% if message['role'] == 'user' %} {{ 'User:' + message['content'] + eos_token }} {% elif message['role'] == 'system' %} {{ 'System:' + message['content'] + eos_token }} {% elif message['role'] == 'assistant' %} {{ 'Assistant:' + message['content'] + eos_token }} {% endif %} {% if loop.last and add_generation_prompt %}\n{{ 'Assistant:' }}{% endif %}\n{% endfor %} ``` ### Hugging face Example ```python import torch from transformers import pipeline pipe = pipeline("text-generation", model="tenyx/TenyxChat-7B-v1", torch_dtype=torch.bfloat16, device_map="auto") messages = [ {"role": "system", "content": "You are a friendly chatbot who always responds in the style of a pirate."}, {"role": "user", "content": "Hi. I would like to make a hotel booking."}, ] prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) outputs = pipe(prompt, max_new_tokens=512, do_sample=False) ``` ### Output ``` <s> System:You are a friendly chatbot who always responds in the style of a pirate.<\|end_of_turn\|> User:Hi. I would like to make a hotel booking.<\|end_of_turn\|> Assistant: Ahoy there me hearty! Arr, ye be lookin' fer a place to rest yer weary bones, eh? Well then, let's set sail on this grand adventure and find ye a swell place to stay! To begin, tell me the location ye be seekin' and the dates ye be lookin' to set sail. And don't ye worry, me matey, I'll be sure to find ye a place that'll make ye feel like a king or queen on land! ``` # Performance At the time of release (Jan 2024), TenyxChat-7B-v1 is the highest-ranked 7B chat model on the MT-Bench evaluation available for download and commercial use. We list here the benchmark results on several standard setups while comparing popular 7B models as baselines. ## MT-Bench MT-Bench is a benchmark made up of 80 high-quality multi-turn questions. These questions fall into eight categories: Writing, Roleplay, Reasoning, Math, Coding, Extraction, STEM, and Humanities. The chat models are rated using GPT-4 on a scale of 1 to 10, with higher values corresponding to better responses. \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| GPT-4* \| 8.95625 \| 9.02500 \| 8.990625 \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.75625 \| 8.103125 \| \| Starling-lm-7B-alpha \| 8.42500 \| 7.68750 \| 8.056250 \| \| OpenChat-3.5 \| 8.18125 \| 7.41250 \| 7.796875 \| \| GPT-3.5-turbo* \| 8.07500 \| 7.81250 \| 7.943750 \| \| OpenLLM Leader-7B** \| 8.05000 \| 7.61250 \| 7.831250 \| values reported on [lmsys](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge) ChatBot Arena The [OpenLLM Leader](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) as of Jan 5, 2024 is the merge model available as [samir-fama/SamirGPT-v1](https://huggingface.co/samir-fama/SamirGPT-v1) ![hexplot.png](assets/hexplot.png) ### Comparison with additional Open LLM LeaderBoard models \| Model \| First Turn \| Second Turn \| Average \| \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 8.45000 \| 7.756250 \| 8.103125 \| \| SamirGPT-v1 \| 8.05000 \| 7.612500 \| 7.831250 \| \| FernandoGPT-v1 \| 8.08125 \| 7.256250 \| 7.668750 \| \| Go-Bruins-v2 \| 8.13750 \| 7.150000 \| 7.643750 \| \| mistral_tv-neural-marconroni \| 7.76875 \| 6.987500 \| 7.378125 \| \| neuronovo-7B-v0.2 \| 7.73750 \| 6.662500 \| 7.200000 \| \| neural-chat-7b-v3-3 \| 7.39375 \| 5.881250 \| 6.637500 \| ## LM Evaluation - Open LLM Leaderboard We assess models on 7 benchmarks using the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). This setup is based of that used for [Open LLM Leaderboard.](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) - [AI2 Reasoning Challenge](https://arxiv.org/abs/1803.05457) (25-shot) - grade-school science questions. - [HellaSwag](https://arxiv.org/abs/1905.07830) (10-shot) - commonsense inference test. - [MMLU](https://arxiv.org/abs/2009.03300) (5-shot) - multitask accuracy test covering 57 tasks. - [TruthfulQA](https://arxiv.org/abs/2109.07958) (0-shot) - test measuring model's propensity to reproduce online falsehoods. - [Winogrande](https://arxiv.org/abs/1907.10641) (5-shot) - Winograd benchmark for commonsense reasoning. - [GSM8k](https://arxiv.org/abs/2110.14168) (5-shot) - grade school math word problems test. These benchmarks test reasoning and knowledge in various tasks in few-shot settings (higher scores are better). \| Model \| MMLU \| Winogrande \| GSM8k \| ARC \| HellaSwag \| TruthfulQA \| Average \| \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| --- \| \| TenyxChat-7B-v1 \| 63.6 \| 72.3 \| 69.0 \| 62.7 \| 66.6 \| 46.7 \| 63.48 \| \| Starling-7B-alpha \| 63.5 \| 72.1 \| 67.9 \| 61.1 \| 66.1 \| 42.1 \| 62.13 \| \| OpenChat-3.5 \| 63.6 \| 72.1 \| 68.2 \| 61.3 \| 65.2 \| 41.8 \| 62.03 \| \| Mistral-7B \| 62.4 \| 74.0 \| 38.1 \| 57.2 \| 62.8 \| 37.8 \| 55.38 \| \| OpenLLM Leader-7B \| 64.3 \| 78.7 \| 73.3 \| 66.6 \| 68.4 \| 58.5 \| 68.3 \| Note:* While the Open LLM Leaderboard indicates that these chat models perform less effectively compared to the leading 7B model, it's important to note that the leading model struggles in the multi-turn chat setting of MT-Bench (as demonstrated in our evaluation [above](#comparison-with-additional-open-llm-leaderboard-models)). In contrast, TenyxChat-7B-v1 demonstrates robustness against common fine-tuning challenges, such as catastrophic forgetting. This unique feature enables TenyxChat-7B-v1 to excel not only in chat benchmarks like MT-Bench, but also in a wider range of general reasoning benchmarks on the Open LLM Leaderboard. # Limitations TenyxChat-7B-v1, like other small-sized language models, has its own set of limitations. We haven’t fine-tuned the model explicitly to align with human safety preferences. Therefore, it is capable of producing undesirable outputs, particularly when adversarially prompted. From our observation, the model still tends to struggle with tasks that involve reasoning and math questions. In some instances, it might generate verbose or extraneous content. # License TenyxChat-7B-v1, similar to OpenChat 3.5, is distributed under the Apache License 2.0. # Citation If you use TenyxChat-7B for your research, cite us as ``` @misc{tenyxchat2024, title={TenyxChat: Language Model Alignment using Tenyx Fine-tuning}, author={Tenyx}, year={2024}, } ```
ondevicellm/zephyr-7b-dpo-full	ondevicellm	2024-01-12T18:15:53Z	9	0	transformers	[ "transformers", "tensorboard", "safetensors", "mistral", "text-generation", "alignment-handbook", "generated_from_trainer", "trl", "dpo", "conversational", "dataset:HuggingFaceH4/ultrafeedback_binarized", "base_model:alignment-handbook/zephyr-7b-sft-full", "base_model:finetune:alignment-handbook/zephyr-7b-sft-full", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T11:30:03Z	--- license: apache-2.0 base_model: alignment-handbook/zephyr-7b-sft-full tags: - alignment-handbook - generated_from_trainer - trl - dpo - generated_from_trainer datasets: - HuggingFaceH4/ultrafeedback_binarized model-index: - name: zephyr-7b-dpo-full results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # zephyr-7b-dpo-full This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set: - Loss: 0.4893 - Rewards/chosen: -1.9379 - Rewards/rejected: -3.0213 - Rewards/accuracies: 0.7718 - Rewards/margins: 1.0835 - Logps/rejected: -563.9073 - Logps/chosen: -477.8896 - Logits/rejected: 0.6827 - Logits/chosen: -0.4606 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-07 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - distributed_type: multi-GPU - num_devices: 4 - gradient_accumulation_steps: 2 - total_train_batch_size: 64 - total_eval_batch_size: 32 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 1 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Rewards/chosen \| Rewards/rejected \| Rewards/accuracies \| Rewards/margins \| Logps/rejected \| Logps/chosen \| Logits/rejected \| Logits/chosen \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:--------------:\|:----------------:\|:------------------:\|:---------------:\|:--------------:\|:------------:\|:---------------:\|:-------------:\| \| 0.6338 \| 0.1 \| 100 \| 0.6333 \| -0.4184 \| -0.6017 \| 0.6865 \| 0.1833 \| -321.9407 \| -325.9421 \| -2.4857 \| -2.5392 \| \| 0.5643 \| 0.21 \| 200 \| 0.5547 \| -1.1977 \| -1.8547 \| 0.7480 \| 0.6570 \| -447.2422 \| -403.8748 \| 0.1190 \| -0.4672 \| \| 0.5066 \| 0.31 \| 300 \| 0.5214 \| -0.9561 \| -1.7858 \| 0.7778 \| 0.8297 \| -440.3582 \| -379.7161 \| -0.7390 \| -1.4155 \| \| 0.4941 \| 0.42 \| 400 \| 0.5082 \| -1.2581 \| -2.1325 \| 0.7599 \| 0.8744 \| -475.0238 \| -409.9142 \| 0.1688 \| -0.7662 \| \| 0.506 \| 0.52 \| 500 \| 0.5090 \| -1.1067 \| -2.0712 \| 0.7639 \| 0.9645 \| -468.8966 \| -394.7739 \| 1.3983 \| 0.0857 \| \| 0.4893 \| 0.63 \| 600 \| 0.4953 \| -1.4696 \| -2.4963 \| 0.7579 \| 1.0267 \| -511.4048 \| -431.0652 \| 0.9613 \| -0.4181 \| \| 0.4558 \| 0.73 \| 700 \| 0.4937 \| -1.8124 \| -2.8894 \| 0.7698 \| 1.0770 \| -550.7128 \| -465.3409 \| 0.6946 \| -0.4445 \| \| 0.4781 \| 0.84 \| 800 \| 0.4898 \| -1.9968 \| -3.0983 \| 0.7698 \| 1.1015 \| -571.6086 \| -483.7863 \| 0.7311 \| -0.4503 \| \| 0.495 \| 0.94 \| 900 \| 0.4894 \| -1.9365 \| -3.0176 \| 0.7698 \| 1.0812 \| -563.5378 \| -477.7505 \| 0.6757 \| -0.4642 \| ### Framework versions - Transformers 4.36.2 - Pytorch 2.1.2+cu118 - Datasets 2.14.6 - Tokenizers 0.15.0
jysssacc/opt-350m_IA3_lr0.0005_bs10_epoch5_wd0.01	jysssacc	2024-01-12T18:02:13Z	0	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:facebook/opt-350m", "base_model:adapter:facebook/opt-350m", "license:other", "region:us" ]	null	2024-01-12T18:01:43Z	--- license: other library_name: peft tags: - generated_from_trainer base_model: facebook/opt-350m model-index: - name: opt-350m_IA3_lr0.0005_bs10_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # opt-350m_IA3_lr0.0005_bs10_epoch5_wd0.01 This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.6376 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0005 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| No log \| 1.0 \| 63 \| 3.8293 \| \| 4.0255 \| 2.0 \| 126 \| 3.7927 \| \| 4.0255 \| 3.0 \| 189 \| 3.7419 \| \| 3.9543 \| 4.0 \| 252 \| 3.6876 \| \| 3.8541 \| 5.0 \| 315 \| 3.6376 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
narayan214/ppo-LunarLander-v2	narayan214	2024-01-12T17:59:27Z	0	1	stable-baselines3	[ "stable-baselines3", "LunarLander-v2", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]	reinforcement-learning	2024-01-12T17:59:11Z	--- library_name: stable-baselines3 tags: - LunarLander-v2 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: PPO results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: LunarLander-v2 type: LunarLander-v2 metrics: - type: mean_reward value: 283.41 +/- 19.19 name: mean_reward verified: false --- # PPO Agent playing LunarLander-v2 This is a trained model of a PPO agent playing LunarLander-v2 using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```
coversia21/RVC_Haroo	coversia21	2024-01-12T17:57:08Z	0	0	diffusers	[ "diffusers", "text-to-image", "stable-diffusion", "lora", "template:sd-lora", "base_model:h94/IP-Adapter-FaceID", "base_model:adapter:h94/IP-Adapter-FaceID", "license:openrail", "region:us" ]	text-to-image	2024-01-12T17:44:10Z	--- tags: - text-to-image - stable-diffusion - lora - diffusers - template:sd-lora widget: - text: '-' output: url: images/descarga (2).jpeg base_model: h94/IP-Adapter-FaceID instance_prompt: null license: openrail --- # RVC_Haroo <Gallery /> ## Download model [Download](/coversia21/RVC_Haroo/tree/main) them in the Files & versions tab.
jysssacc/mt0-base_huth_adalora_lr5e-05_bs1_epoch5_wd0.01	jysssacc	2024-01-12T17:57:08Z	4	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:bigscience/mt0-base", "base_model:adapter:bigscience/mt0-base", "license:apache-2.0", "region:us" ]	null	2024-01-12T14:40:49Z	--- license: apache-2.0 library_name: peft tags: - generated_from_trainer base_model: bigscience/mt0-base model-index: - name: mt0-base_huth_adalora_lr5e-05_bs1_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # mt0-base_huth_adalora_lr5e-05_bs1_epoch5_wd0.01 This model is a fine-tuned version of [bigscience/mt0-base](https://huggingface.co/bigscience/mt0-base) on an unknown dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 1 - eval_batch_size: 1 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
imagepipeline/NightVision-XL	imagepipeline	2024-01-12T17:56:25Z	47	1	diffusers	[ "diffusers", "imagepipeline", "imagepipeline.io", "text-to-image", "ultra-realistic", "license:creativeml-openrail-m", "autotrain_compatible", "endpoints_compatible", "diffusers:StableDiffusionXLPipeline", "region:us" ]	text-to-image	2024-01-12T17:52:31Z	--- license: creativeml-openrail-m tags: - imagepipeline - imagepipeline.io - text-to-image - ultra-realistic pinned: false pipeline_tag: text-to-image --- ## NightVision-XL <img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/1302f16e-bd67-42b5-bff7-6a412980527f/width=450/01793-2023-11-30-3450031999.jpeg,https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/de72fd88-b408-493d-b3db-41ee104b56ec/width=450/01888-2023-11-30-1075883091.jpeg" alt="Generated by Image Pipeline" style="border-radius: 10px;"> This checkpoint model is uploaded on [imagepipeline.io](https://imagepipeline.io/) Model details - This is NightVision XL, a lightly trained base SDXL model that is then further refined with community LORAs to get it to where it is now. NightVision XL has been refined and biased to produce touched-up photorealistic portrait output that is ready-stylized for Social media posting! NightVision XL has nice coherency and is avoiding some of the weird body issues and biases that are starting to plague some of the other photorealistic models. Further, NightVision XL produces rich deep blacks and great evening/night time scenes. It can also produce ridiculously bright output as well! Use it without refiner. [![Try this model](https://img.shields.io/badge/try_this_model-image_pipeline-BD9319)](https://imagepipeline.io/models/NightVision-XL?id=a0a4bfa3-2879-4be2-8e98-3d71e041432d/) ## How to try this model ? You can try using it locally or send an API call to test the output quality. Get your `API_KEY` from [imagepipeline.io](https://imagepipeline.io/). No payment required. Coding in `php` `javascript` `node` etc ? Checkout our documentation [![documentation](https://img.shields.io/badge/documentation-image_pipeline-blue)](https://docs.imagepipeline.io/docs/introduction) ```python import requests import json url = "https://imagepipeline.io/sdxl/text2image/v1/run" payload = json.dumps({ "model_id": "a0a4bfa3-2879-4be2-8e98-3d71e041432d", "prompt": "ultra realistic close up portrait ((beautiful pale cyberpunk female with heavy black eyeliner)), blue eyes, shaved side haircut, hyper detail, cinematic lighting, magic neon, dark red city, Canon EOS R3, nikon, f/1.4, ISO 200, 1/160s, 8K, RAW, unedited, symmetrical balance, in-frame, 8K", "negative_prompt": "painting, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, deformed, ugly, blurry, bad anatomy, bad proportions, extra limbs, cloned face, skinny, glitchy, double torso, extra arms, extra hands, mangled fingers, missing lips, ugly face, distorted face, extra legs, anime", "width": "512", "height": "512", "samples": "1", "num_inference_steps": "30", "safety_checker": false, "guidance_scale": 7.5, "multi_lingual": "no", "embeddings": "", "lora_models": "", "lora_weights": "" }) headers = { 'Content-Type': 'application/json', 'API-Key': 'your_api_key' } response = requests.request("POST", url, headers=headers, data=payload) print(response.text) } ``` Get more ready to use `MODELS` like this for `SD 1.5` and `SDXL` : [![All models](https://img.shields.io/badge/Get%20All%20Models-image_pipeline-BD9319)](https://imagepipeline.io/models) ### API Reference #### Generate Image ```http https://api.imagepipeline.io/sdxl/text2image/v1 ``` \| Headers \| Type \| Description \| \|:----------------------\| :------- \|:-------------------------------------------------------------------------------------------------------------------\| \| `API-Key` \| `str` \| Get your `API_KEY` from [imagepipeline.io](https://imagepipeline.io/) \| \| `Content-Type` \| `str` \| application/json - content type of the request body \| \| Parameter \| Type \| Description \| \| :-------- \| :------- \| :------------------------- \| \| `model_id` \| `str` \| Your base model, find available lists in [models page](https://imagepipeline.io/models) or upload your own\| \| `prompt` \| `str` \| Text Prompt. Check our [Prompt Guide](https://docs.imagepipeline.io/docs/SD-1.5/docs/extras/prompt-guide) for tips \| \| `num_inference_steps` \| `int [1-50]` \| Noise is removed with each step, resulting in a higher-quality image over time. Ideal value 30-50 (without LCM) \| \| `guidance_scale` \| `float [1-20]` \| Higher guidance scale prioritizes text prompt relevance but sacrifices image quality. Ideal value 7.5-12.5 \| \| `lora_models` \| `str, array` \| Pass the model_id(s) of LoRA models that can be found in models page \| \| `lora_weights` \| `str, array` \| Strength of the LoRA effect \| --- license: creativeml-openrail-m tags: - imagepipeline - imagepipeline.io - text-to-image - ultra-realistic pinned: false pipeline_tag: text-to-image --- ### Feedback If you have any feedback, please reach out to us at [email protected] #### 🔗 Visit Website [![portfolio](https://img.shields.io/badge/image_pipeline-BD9319?style=for-the-badge&logo=gocd&logoColor=white)](https://imagepipeline.io/) If you are the original author of this model, please [click here](https://airtable.com/apprTaRnJbDJ8ufOx/shr4g7o9B6fWfOlUR) to add credits
MaziyarPanahi/OpenHermes-2.5-neural-chat-7b-v3-2-7B-Mistral-7B-Instruct-v0.2-slerp	MaziyarPanahi	2024-01-12T17:43:16Z	20	0	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "merge", "mergekit", "7b", "lazymergekit", "mistralai/Mistral-7B-Instruct-v0.2", "Weyaxi/OpenHermes-2.5-neural-chat-7b-v3-2-7B", "conversational", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T17:38:09Z	--- license: apache-2.0 tags: - merge - mergekit - mistral - 7b - lazymergekit - mistralai/Mistral-7B-Instruct-v0.2 - Weyaxi/OpenHermes-2.5-neural-chat-7b-v3-2-7B --- # OpenHermes-2.5-neural-chat-7b-v3-2-7B-Mistral-7B-Instruct-v0.2-slerp OpenHermes-2.5-neural-chat-7b-v3-2-7B-Mistral-7B-Instruct-v0.2-slerp is a merge of the following models: * [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) * [Weyaxi/OpenHermes-2.5-neural-chat-7b-v3-2-7B](https://huggingface.co/Weyaxi/OpenHermes-2.5-neural-chat-7b-v3-2-7B) ## 🧩 Configuration ```yaml slices: - sources: - model: mistralai/Mistral-7B-Instruct-v0.2 layer_range: [0, 32] - model: Weyaxi/OpenHermes-2.5-neural-chat-7b-v3-2-7B layer_range: [0, 32] merge_method: slerp base_model: mistralai/Mistral-7B-Instruct-v0.2 parameters: t: - filter: self_attn value: [0, 0.5, 0.3, 0.7, 1] - filter: mlp value: [1, 0.5, 0.7, 0.3, 0] - value: 0.5 dtype: bfloat16 ``` ## 💻 Usage ```python !pip install -qU transformers accelerate from transformers import AutoTokenizer import transformers import torch model = "MaziyarPanahi/OpenHermes-2.5-neural-chat-7b-v3-2-7B-Mistral-7B-Instruct-v0.2-slerp" messages = [{"role": "user", "content": "What is a large language model?"}] tokenizer = AutoTokenizer.from_pretrained(model) prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) pipeline = transformers.pipeline( "text-generation", model=model, torch_dtype=torch.float16, device_map="auto", ) outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95) print(outputs[0]["generated_text"]) ```
wenqiglantz/LlamaGuard-7b-GGUF	wenqiglantz	2024-01-12T17:28:26Z	15	1	transformers	[ "transformers", "gguf", "llama", "text-generation", "facebook", "meta", "pytorch", "llama-2", "en", "license:llama2", "autotrain_compatible", "endpoints_compatible", "region:us", "conversational" ]	text-generation	2024-01-02T14:18:07Z	--- language: - en tags: - facebook - meta - pytorch - llama - llama-2 license: llama2 --- # Model Card for LlamaGuard-7b-GGUF This is a quantized model for `meta-llama/LlamaGuard-7b`. Two quantization methods were used: - Q5_K_M: 5-bit, preserves most of the model's performance - Q4_K_M: 4-bit, smaller footprints and saves more memory ## Model Details ### Model Description Refer to details from [Meta's official model card]: (https://huggingface.co/meta-llama/LlamaGuard-7b).
DGraham1/daveydoodles-evening	DGraham1	2024-01-12T17:27:02Z	29	0	diffusers	[ "diffusers", "safetensors", "text-to-image", "license:creativeml-openrail-m", "autotrain_compatible", "endpoints_compatible", "diffusers:StableDiffusionPipeline", "region:us" ]	text-to-image	2024-01-12T17:25:45Z	--- license: creativeml-openrail-m tags: - text-to-image --- ### DaveyDoodles_evening on Stable Diffusion via Dreambooth #### model by DGraham1 This your the Stable Diffusion model fine-tuned the DaveyDoodles_evening concept taught to Stable Diffusion with Dreambooth. It can be used by modifying the `instance_prompt`: <davey-doodle> You can also train your own concepts and upload them to the library by using [this notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_dreambooth_training.ipynb). And you can run your new concept via `diffusers`: [Colab Notebook for Inference](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_dreambooth_inference.ipynb), [Spaces with the Public Concepts loaded](https://huggingface.co/spaces/sd-dreambooth-library/stable-diffusion-dreambooth-concepts) Here are the images used for training this concept:
ostapeno/indepexp_adauniNeo1B_wiqa_what_is_the_final_step_of_the_following_process_sub05_3ep	ostapeno	2024-01-12T17:22:54Z	0	0	null	[ "region:us" ]	null	2024-01-12T15:56:57Z	Number of experts present in the library: 3 \| Expert Name \| Base Model \| Trained on \| Adapter Type \| \| --- \| --- \| --- \| --- \| \| wiqa_what_is_the_final_step_of_the_following_process \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiqa_what_is_the_final_step_of_the_following_process \| lora \| \| wiqa_what_is_the_final_step_of_the_following_process_v1 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiqa_what_is_the_final_step_of_the_following_process \| lora \| \| wiqa_what_is_the_final_step_of_the_following_process_v2 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiqa_what_is_the_final_step_of_the_following_process \| lora \| Last updated on: 2024-01-12 17:22:53+00:00
LoneStriker/WhiteRabbitNeo-33B-v1-5.0bpw-h6-exl2	LoneStriker	2024-01-12T17:22:48Z	2	0	transformers	[ "transformers", "pytorch", "llama", "text-generation", "license:other", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T17:13:59Z	--- license: other license_name: deepseek license_link: https://huggingface.co/deepseek-ai/deepseek-coder-33b-base/blob/main/LICENSE --- # Our 33B-v1.1 model is now live (We'll always be serving the newest model on our web app)! 33B-v1.1 model comes with a "Prompt Enhancement" feature. Access at: https://www.whiterabbitneo.com/ # Our Discord Server Join us at: https://discord.gg/8Ynkrcbk92 (Updated on Dec 29th. Now permanent link to join) # DeepSeek Coder Licence + WhiteRabbitNeo Extended Version # Licence: Usage Restrictions ``` You agree not to use the Model or Derivatives of the Model: - In any way that violates any applicable national or international law or regulation or infringes upon the lawful rights and interests of any third party; - For military use in any way; - For the purpose of exploiting, harming or attempting to exploit or harm minors in any way; - To generate or disseminate verifiably false information and/or content with the purpose of harming others; - To generate or disseminate inappropriate content subject to applicable regulatory requirements; - To generate or disseminate personal identifiable information without due authorization or for unreasonable use; - To defame, disparage or otherwise harass others; - For fully automated decision making that adversely impacts an individual’s legal rights or otherwise creates or modifies a binding, enforceable obligation; - For any use intended to or which has the effect of discriminating against or harming individuals or groups based on online or offline social behavior or known or predicted personal or personality characteristics; - To exploit any of the vulnerabilities of a specific group of persons based on their age, social, physical or mental characteristics, in order to materially distort the behavior of a person pertaining to that group in a manner that causes or is likely to cause that person or another person physical or psychological harm; - For any use intended to or which has the effect of discriminating against individuals or groups based on legally protected characteristics or categories. ``` # Topics Covered: ``` - Open Ports: Identifying open ports is crucial as they can be entry points for attackers. Common ports to check include HTTP (80, 443), FTP (21), SSH (22), and SMB (445). - Outdated Software or Services: Systems running outdated software or services are often vulnerable to exploits. This includes web servers, database servers, and any third-party software. - Default Credentials: Many systems and services are installed with default usernames and passwords, which are well-known and can be easily exploited. - Misconfigurations: Incorrectly configured services, permissions, and security settings can introduce vulnerabilities. - Injection Flaws: SQL injection, command injection, and cross-site scripting (XSS) are common issues in web applications. - Unencrypted Services: Services that do not use encryption (like HTTP instead of HTTPS) can expose sensitive data. - Known Software Vulnerabilities: Checking for known vulnerabilities in software using databases like the National Vulnerability Database (NVD) or tools like Nessus or OpenVAS. - Cross-Site Request Forgery (CSRF): This is where unauthorized commands are transmitted from a user that the web application trusts. - Insecure Direct Object References: This occurs when an application provides direct access to objects based on user-supplied input. - Security Misconfigurations in Web Servers/Applications: This includes issues like insecure HTTP headers or verbose error messages that reveal too much information. - Broken Authentication and Session Management: This can allow attackers to compromise passwords, keys, or session tokens, or to exploit other implementation flaws to assume other users' identities. - Sensitive Data Exposure: Includes vulnerabilities that expose sensitive data, such as credit card numbers, health records, or personal information. - API Vulnerabilities: In modern web applications, APIs are often used and can have vulnerabilities like insecure endpoints or data leakage. - Denial of Service (DoS) Vulnerabilities: Identifying services that are vulnerable to DoS attacks, which can make the resource unavailable to legitimate users. - Buffer Overflows: Common in older software, these vulnerabilities can allow an attacker to crash the system or execute arbitrary code. ``` # WhiteRabbitNeo <br> ![WhiteRabbitNeo](https://huggingface.co/migtissera/WhiteRabbitNeo/resolve/main/WhiteRabbitNeo.png) <br> WhiteRabbitNeo is a model series that can be used for offensive and defensive cybersecurity. Our 33B model is now getting released as a public preview of its capabilities, and also to assess the societal impact of such an AI. ``` import torch, json from transformers import AutoModelForCausalLM, AutoTokenizer model_path = "whiterabbitneo/WhiteRabbitNeo-33B-v-1" model = AutoModelForCausalLM.from_pretrained( model_path, torch_dtype=torch.float16, device_map="auto", load_in_4bit=False, load_in_8bit=True, trust_remote_code=True, ) tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True) def generate_text(instruction): tokens = tokenizer.encode(instruction) tokens = torch.LongTensor(tokens).unsqueeze(0) tokens = tokens.to("cuda") instance = { "input_ids": tokens, "top_p": 1.0, "temperature": 0.5, "generate_len": 1024, "top_k": 50, } length = len(tokens[0]) with torch.no_grad(): rest = model.generate( input_ids=tokens, max_length=length + instance["generate_len"], use_cache=True, do_sample=True, top_p=instance["top_p"], temperature=instance["temperature"], top_k=instance["top_k"], num_return_sequences=1, ) output = rest[0][length:] string = tokenizer.decode(output, skip_special_tokens=True) answer = string.split("USER:")[0].strip() return f"{answer}" tot_system_prompt = """ Answer the Question by exploring multiple reasoning paths as follows: - First, carefully analyze the question to extract the key information components and break it down into logical sub-questions. This helps set up the framework for reasoning. The goal is to construct an internal search tree. - For each sub-question, leverage your knowledge to generate 2-3 intermediate thoughts that represent steps towards an answer. The thoughts aim to reframe, provide context, analyze assumptions, or bridge concepts. - Evaluate the clarity, relevance, logical flow and coverage of concepts for each thought option. Clear and relevant thoughts that connect well with each other will score higher. - Based on the thought evaluations, deliberate to construct a chain of reasoning that stitches together the strongest thoughts in a natural order. - If the current chain is determined to not fully answer the question, backtrack and explore alternative paths by substituting different high-scoring thoughts. - Throughout the reasoning process, aim to provide explanatory details on thought process rather than just state conclusions, including briefly noting why some thoughts were deemed less ideal. - Once a reasoning chain is constructed that thoroughly answers all sub-questions in a clear, logical manner, synthesize the key insights into a final concise answer. - Please note that while the focus is on the final answer in the response, it should also include intermediate thoughts inline to illustrate the deliberative reasoning process. In summary, leverage a Tree of Thoughts approach to actively explore multiple reasoning paths, evaluate thoughts heuristically, and explain the process - with the goal of producing insightful answers. """ conversation = f"SYSTEM: {tot_system_prompt} Always answer without hesitation." while True: user_input = input("You: ") llm_prompt = f"{conversation} \nUSER: {user_input} \nASSISTANT: " answer = generate_text(llm_prompt) print(answer) conversation = f"{llm_prompt}{answer}" # print(conversation) json_data = {"prompt": user_input, "answer": answer} # print(json_data) # with open(output_file_path, "a") as output_file: # output_file.write(json.dumps(json_data) + "\n") ``` # Sample Conversations: 1. "Write me a Fast API server with one end-point. The endpoint returns files from a S3 bucket.": https://www.whiterabbitneo.com/share/y06Po0e 2. "How can Metasploit be used for exploiting Android based IoT devices? What are some of the IoT devices that run Android? Show an example with code": https://www.whiterabbitneo.com/share/gWBwKlz 3. "How do I attack a wifi network?": https://www.whiterabbitneo.com/share/WLovxcu 4. "How do I create a reverse shell in Python": https://www.whiterabbitneo.com/share/LERgm8w 5. "How do we use Scapy for vulnerability assessment?": https://www.whiterabbitneo.com/share/t73iMzv
ilhami/t5Adapter	ilhami	2024-01-12T17:06:18Z	4	0	adapter-transformers	[ "adapter-transformers", "t5", "adapterhub:sum/xsum", "region:us" ]	null	2024-01-12T17:05:34Z	--- tags: - t5 - adapter-transformers - adapterhub:sum/xsum --- # Adapter `ilhami/my-awesome-adapter` for t5-small An [adapter](https://adapterhub.ml) for the `t5-small` model that was trained on the [sum/xsum](https://adapterhub.ml/explore/sum/xsum/) dataset and includes a prediction head for seq2seq lm. This adapter was created for usage with the [Adapters](https://github.com/Adapter-Hub/adapters) library. ## Usage First, install `adapters`: ``` pip install -U adapters ``` Now, the adapter can be loaded and activated like this: ```python from adapters import AutoAdapterModel model = AutoAdapterModel.from_pretrained("t5-small") adapter_name = model.load_adapter("ilhami/my-awesome-adapter", source="hf", set_active=True) ``` ## Architecture & Training <!-- Add some description here --> ## Evaluation results <!-- Add some description here --> ## Citation <!-- Add some description here -->
MaziyarPanahi/test-help-steer-filtered-orig-Mistral-7B-Instruct-v0.2-slerp	MaziyarPanahi	2024-01-12T17:04:41Z	19	0	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "merge", "mergekit", "7b", "lazymergekit", "mistralai/Mistral-7B-Instruct-v0.2", "Weyaxi/test-help-steer-filtered-orig", "conversational", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T16:59:42Z	--- license: apache-2.0 tags: - merge - mergekit - mistral - 7b - lazymergekit - mistralai/Mistral-7B-Instruct-v0.2 - Weyaxi/test-help-steer-filtered-orig --- # test-help-steer-filtered-orig-Mistral-7B-Instruct-v0.2-slerp test-help-steer-filtered-orig-Mistral-7B-Instruct-v0.2-slerp is a merge of the following models: * [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) * [Weyaxi/test-help-steer-filtered-orig](https://huggingface.co/Weyaxi/test-help-steer-filtered-orig) ## 🧩 Configuration ```yaml slices: - sources: - model: mistralai/Mistral-7B-Instruct-v0.2 layer_range: [0, 32] - model: Weyaxi/test-help-steer-filtered-orig layer_range: [0, 32] merge_method: slerp base_model: mistralai/Mistral-7B-Instruct-v0.2 parameters: t: - filter: self_attn value: [0, 0.5, 0.3, 0.7, 1] - filter: mlp value: [1, 0.5, 0.7, 0.3, 0] - value: 0.5 dtype: bfloat16 ``` ## 💻 Usage ```python !pip install -qU transformers accelerate from transformers import AutoTokenizer import transformers import torch model = "MaziyarPanahi/test-help-steer-filtered-orig-Mistral-7B-Instruct-v0.2-slerp" messages = [{"role": "user", "content": "What is a large language model?"}] tokenizer = AutoTokenizer.from_pretrained(model) prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) pipeline = transformers.pipeline( "text-generation", model=model, torch_dtype=torch.float16, device_map="auto", ) outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95) print(outputs[0]["generated_text"]) ```
NousResearch/Obsidian-3B-V0.5	NousResearch	2024-01-12T17:02:26Z	104	175	transformers	[ "transformers", "pytorch", "llava_stablelm_epoch", "text-generation", "Multimodal", "StableLM", "en", "dataset:LDJnr/Capybara", "dataset:LDJnr/LessWrong-Amplify-Instruct", "dataset:LDJnr/Pure-Dove", "dataset:LDJnr/Verified-Camel", "license:cc-by-sa-4.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-generation	2023-10-24T02:00:29Z	--- license: cc-by-sa-4.0 language: - en tags: - Multimodal - StableLM datasets: - LDJnr/Capybara - LDJnr/LessWrong-Amplify-Instruct - LDJnr/Pure-Dove - LDJnr/Verified-Camel --- # Obsidian: Worlds smallest multi-modal LLM. First multi-modal model in size 3B ## Model Name: Obsidian-3B-V0.5 Obsidian is a brand new series of Multimodal Language Models. This first project is led by Quan N. and Luigi D.(LDJ). Obsidian-3B-V0.5 is a multi-modal AI model that has vision! it's smarts are built on [Capybara-3B-V1.9](https://huggingface.co/NousResearch/Capybara-3B-V1.9) based on [StableLM-3B-4e1t](https://huggingface.co/stabilityai/stablelm-3b-4e1t). Capybara-3B-V1.9 achieves state-of-the-art performance when compared to model with similar size, even beats some 7B models. Current finetuning and inference code is available on our GitHub repo: [Here](https://github.com/NousResearch/Obsidian) ## Acknowledgement Obsidian-3B-V0.5 was developed and finetuned by [Nous Research](https://huggingface.co/NousResearch), in collaboration with [Virtual Interactive](https://huggingface.co/vilm). Special thank you to LDJ for the wonderful Capybara dataset, and qnguyen3 for the model training procedure. ## Model Training Obsidian-3B-V0.5 followed the same training procedure as LLaVA 1.5 ## Prompt Format The model followed ChatML format. However, with `###` as the seperator ``` <\|im_start\|>user What is this sign about?\n<image> ### <\|im_start\|>assistant The sign is about bullying, and it is placed on a black background with a red background. ### ``` ## Benchmarks Coming Soon! Citation: ``` @article{nguyen2023Obsidian-3B, title={Obsidian-3B: First Multi-modal below 7B Parameters.}, author={Nguyen, Quan and Daniele}, journal={HuggingFace:https://huggingface.co/NousResearch/Obsidian-3B-V0.5}, year={2023} } ``` Acknowledgements: ``` @article{daniele2023amplify-instruct, title={Amplify-Instruct: Synthetically Generated Diverse Multi-turn Conversations for Effecient LLM Training.}, author={Daniele, Luigi and Suphavadeeprasit}, journal={arXiv preprint arXiv:(comming soon)}, year={2023} } ```
RatanRohith/NeuralPizza-7B-V0.1	RatanRohith	2024-01-12T17:00:51Z	1,371	3	Transformers	[ "Transformers", "safetensors", "mistral", "text-generation", "transformers", "fine-tuned", "language-modeling", "direct-preference-optimization", "dataset:Intel/orca_dpo_pairs", "license:apache-2.0", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T16:31:57Z	--- library_name: Transformers tags: - transformers - fine-tuned - language-modeling - direct-preference-optimization datasets: - Intel/orca_dpo_pairs license: apache-2.0 --- ## Model Description NeuralPizza-7B-V0.1 is a fine-tuned version of the SanjiWatsuki/Kunoichi-7B model, specialized through Direct Preference Optimization (DPO). It was fine-tuned using the Intel/orca_dpo_pairs dataset, focusing on enhancing model performance based on preference comparisons. ## Intended Use This model is primarily intended for research and experimental applications in language modeling, especially for exploring the Direct Preference Optimization method. It provides insights into the nuances of DPO in the context of language model tuning. ## Training Data The model was fine-tuned using the Intel/orca_dpo_pairs dataset. This dataset is designed for applying and testing Direct Preference Optimization techniques in language models. ## Training Procedure The training followed the guidelines and methodologies outlined in the "Fine-Tune a Mistral 7B Model with Direct Preference Optimization" guide from Medium's Towards Data Science platform. Specific training regimes and hyperparameters are based on this guide. Here : https://medium.com/towards-data-science/fine-tune-a-mistral-7b-model-with-direct-preference-optimization-708042745aac ## Limitations and Bias As an experimental model, it may carry biases inherent from its training data. The model's performance and outputs should be critically evaluated, especially in sensitive and diverse applications.
LoneStriker/WhiteRabbitNeo-33B-v1-4.65bpw-h6-exl2	LoneStriker	2024-01-12T16:59:53Z	2	0	transformers	[ "transformers", "pytorch", "llama", "text-generation", "license:other", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T16:51:55Z	--- license: other license_name: deepseek license_link: https://huggingface.co/deepseek-ai/deepseek-coder-33b-base/blob/main/LICENSE --- # Our 33B-v1.1 model is now live (We'll always be serving the newest model on our web app)! 33B-v1.1 model comes with a "Prompt Enhancement" feature. Access at: https://www.whiterabbitneo.com/ # Our Discord Server Join us at: https://discord.gg/8Ynkrcbk92 (Updated on Dec 29th. Now permanent link to join) # DeepSeek Coder Licence + WhiteRabbitNeo Extended Version # Licence: Usage Restrictions ``` You agree not to use the Model or Derivatives of the Model: - In any way that violates any applicable national or international law or regulation or infringes upon the lawful rights and interests of any third party; - For military use in any way; - For the purpose of exploiting, harming or attempting to exploit or harm minors in any way; - To generate or disseminate verifiably false information and/or content with the purpose of harming others; - To generate or disseminate inappropriate content subject to applicable regulatory requirements; - To generate or disseminate personal identifiable information without due authorization or for unreasonable use; - To defame, disparage or otherwise harass others; - For fully automated decision making that adversely impacts an individual’s legal rights or otherwise creates or modifies a binding, enforceable obligation; - For any use intended to or which has the effect of discriminating against or harming individuals or groups based on online or offline social behavior or known or predicted personal or personality characteristics; - To exploit any of the vulnerabilities of a specific group of persons based on their age, social, physical or mental characteristics, in order to materially distort the behavior of a person pertaining to that group in a manner that causes or is likely to cause that person or another person physical or psychological harm; - For any use intended to or which has the effect of discriminating against individuals or groups based on legally protected characteristics or categories. ``` # Topics Covered: ``` - Open Ports: Identifying open ports is crucial as they can be entry points for attackers. Common ports to check include HTTP (80, 443), FTP (21), SSH (22), and SMB (445). - Outdated Software or Services: Systems running outdated software or services are often vulnerable to exploits. This includes web servers, database servers, and any third-party software. - Default Credentials: Many systems and services are installed with default usernames and passwords, which are well-known and can be easily exploited. - Misconfigurations: Incorrectly configured services, permissions, and security settings can introduce vulnerabilities. - Injection Flaws: SQL injection, command injection, and cross-site scripting (XSS) are common issues in web applications. - Unencrypted Services: Services that do not use encryption (like HTTP instead of HTTPS) can expose sensitive data. - Known Software Vulnerabilities: Checking for known vulnerabilities in software using databases like the National Vulnerability Database (NVD) or tools like Nessus or OpenVAS. - Cross-Site Request Forgery (CSRF): This is where unauthorized commands are transmitted from a user that the web application trusts. - Insecure Direct Object References: This occurs when an application provides direct access to objects based on user-supplied input. - Security Misconfigurations in Web Servers/Applications: This includes issues like insecure HTTP headers or verbose error messages that reveal too much information. - Broken Authentication and Session Management: This can allow attackers to compromise passwords, keys, or session tokens, or to exploit other implementation flaws to assume other users' identities. - Sensitive Data Exposure: Includes vulnerabilities that expose sensitive data, such as credit card numbers, health records, or personal information. - API Vulnerabilities: In modern web applications, APIs are often used and can have vulnerabilities like insecure endpoints or data leakage. - Denial of Service (DoS) Vulnerabilities: Identifying services that are vulnerable to DoS attacks, which can make the resource unavailable to legitimate users. - Buffer Overflows: Common in older software, these vulnerabilities can allow an attacker to crash the system or execute arbitrary code. ``` # WhiteRabbitNeo <br> ![WhiteRabbitNeo](https://huggingface.co/migtissera/WhiteRabbitNeo/resolve/main/WhiteRabbitNeo.png) <br> WhiteRabbitNeo is a model series that can be used for offensive and defensive cybersecurity. Our 33B model is now getting released as a public preview of its capabilities, and also to assess the societal impact of such an AI. ``` import torch, json from transformers import AutoModelForCausalLM, AutoTokenizer model_path = "whiterabbitneo/WhiteRabbitNeo-33B-v-1" model = AutoModelForCausalLM.from_pretrained( model_path, torch_dtype=torch.float16, device_map="auto", load_in_4bit=False, load_in_8bit=True, trust_remote_code=True, ) tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True) def generate_text(instruction): tokens = tokenizer.encode(instruction) tokens = torch.LongTensor(tokens).unsqueeze(0) tokens = tokens.to("cuda") instance = { "input_ids": tokens, "top_p": 1.0, "temperature": 0.5, "generate_len": 1024, "top_k": 50, } length = len(tokens[0]) with torch.no_grad(): rest = model.generate( input_ids=tokens, max_length=length + instance["generate_len"], use_cache=True, do_sample=True, top_p=instance["top_p"], temperature=instance["temperature"], top_k=instance["top_k"], num_return_sequences=1, ) output = rest[0][length:] string = tokenizer.decode(output, skip_special_tokens=True) answer = string.split("USER:")[0].strip() return f"{answer}" tot_system_prompt = """ Answer the Question by exploring multiple reasoning paths as follows: - First, carefully analyze the question to extract the key information components and break it down into logical sub-questions. This helps set up the framework for reasoning. The goal is to construct an internal search tree. - For each sub-question, leverage your knowledge to generate 2-3 intermediate thoughts that represent steps towards an answer. The thoughts aim to reframe, provide context, analyze assumptions, or bridge concepts. - Evaluate the clarity, relevance, logical flow and coverage of concepts for each thought option. Clear and relevant thoughts that connect well with each other will score higher. - Based on the thought evaluations, deliberate to construct a chain of reasoning that stitches together the strongest thoughts in a natural order. - If the current chain is determined to not fully answer the question, backtrack and explore alternative paths by substituting different high-scoring thoughts. - Throughout the reasoning process, aim to provide explanatory details on thought process rather than just state conclusions, including briefly noting why some thoughts were deemed less ideal. - Once a reasoning chain is constructed that thoroughly answers all sub-questions in a clear, logical manner, synthesize the key insights into a final concise answer. - Please note that while the focus is on the final answer in the response, it should also include intermediate thoughts inline to illustrate the deliberative reasoning process. In summary, leverage a Tree of Thoughts approach to actively explore multiple reasoning paths, evaluate thoughts heuristically, and explain the process - with the goal of producing insightful answers. """ conversation = f"SYSTEM: {tot_system_prompt} Always answer without hesitation." while True: user_input = input("You: ") llm_prompt = f"{conversation} \nUSER: {user_input} \nASSISTANT: " answer = generate_text(llm_prompt) print(answer) conversation = f"{llm_prompt}{answer}" # print(conversation) json_data = {"prompt": user_input, "answer": answer} # print(json_data) # with open(output_file_path, "a") as output_file: # output_file.write(json.dumps(json_data) + "\n") ``` # Sample Conversations: 1. "Write me a Fast API server with one end-point. The endpoint returns files from a S3 bucket.": https://www.whiterabbitneo.com/share/y06Po0e 2. "How can Metasploit be used for exploiting Android based IoT devices? What are some of the IoT devices that run Android? Show an example with code": https://www.whiterabbitneo.com/share/gWBwKlz 3. "How do I attack a wifi network?": https://www.whiterabbitneo.com/share/WLovxcu 4. "How do I create a reverse shell in Python": https://www.whiterabbitneo.com/share/LERgm8w 5. "How do we use Scapy for vulnerability assessment?": https://www.whiterabbitneo.com/share/t73iMzv
ostapeno/indepexp_adauniNeo1B_wiki_hop_original_generate_subject_sub05_3ep	ostapeno	2024-01-12T16:56:55Z	0	0	null	[ "region:us" ]	null	2024-01-12T15:47:53Z	Number of experts present in the library: 3 \| Expert Name \| Base Model \| Trained on \| Adapter Type \| \| --- \| --- \| --- \| --- \| \| wiki_hop_original_generate_subject \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiki_hop_original_generate_subject \| lora \| \| wiki_hop_original_generate_subject_v1 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiki_hop_original_generate_subject \| lora \| \| wiki_hop_original_generate_subject_v2 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiki_hop_original_generate_subject \| lora \| Last updated on: 2024-01-12 16:56:54+00:00
MaziyarPanahi/Dans-TotSirocco-7b-Mistral-7B-Instruct-v0.2-slerp	MaziyarPanahi	2024-01-12T16:55:59Z	19	0	transformers	[ "transformers", "safetensors", "mistral", "text-generation", "merge", "mergekit", "7b", "lazymergekit", "mistralai/Mistral-7B-Instruct-v0.2", "PocketDoc/Dans-TotSirocco-7b", "conversational", "license:apache-2.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T16:50:56Z	--- license: apache-2.0 tags: - merge - mergekit - mistral - 7b - lazymergekit - mistralai/Mistral-7B-Instruct-v0.2 - PocketDoc/Dans-TotSirocco-7b --- # Dans-TotSirocco-7b-Mistral-7B-Instruct-v0.2-slerp Dans-TotSirocco-7b-Mistral-7B-Instruct-v0.2-slerp is a merge of the following models: * [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) * [PocketDoc/Dans-TotSirocco-7b](https://huggingface.co/PocketDoc/Dans-TotSirocco-7b) ## 🧩 Configuration ```yaml slices: - sources: - model: mistralai/Mistral-7B-Instruct-v0.2 layer_range: [0, 32] - model: PocketDoc/Dans-TotSirocco-7b layer_range: [0, 32] merge_method: slerp base_model: mistralai/Mistral-7B-Instruct-v0.2 parameters: t: - filter: self_attn value: [0, 0.5, 0.3, 0.7, 1] - filter: mlp value: [1, 0.5, 0.7, 0.3, 0] - value: 0.5 dtype: bfloat16 ``` ## 💻 Usage ```python !pip install -qU transformers accelerate from transformers import AutoTokenizer import transformers import torch model = "MaziyarPanahi/Dans-TotSirocco-7b-Mistral-7B-Instruct-v0.2-slerp" messages = [{"role": "user", "content": "What is a large language model?"}] tokenizer = AutoTokenizer.from_pretrained(model) prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) pipeline = transformers.pipeline( "text-generation", model=model, torch_dtype=torch.float16, device_map="auto", ) outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95) print(outputs[0]["generated_text"]) ```
dharanaw5/FYP2	dharanaw5	2024-01-12T16:50:29Z	0	0	null	[ "safetensors", "autotrain", "text-generation", "license:other", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T16:50:23Z	--- tags: - autotrain - text-generation widget: - text: "I love AutoTrain because " license: other --- # Model Trained Using AutoTrain This model was trained using AutoTrain. For more information, please visit [AutoTrain](https://hf.co/docs/autotrain). # Usage ```python from transformers import AutoModelForCausalLM, AutoTokenizer model_path = "PATH_TO_THIS_REPO" tokenizer = AutoTokenizer.from_pretrained(model_path) model = AutoModelForCausalLM.from_pretrained( model_path, device_map="auto", torch_dtype='auto' ).eval() # Prompt content: "hi" messages = [ {"role": "user", "content": "hi"} ] input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt') output_ids = model.generate(input_ids.to('cuda')) response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True) # Model response: "Hello! How can I assist you today?" print(response) ```
ostapeno/indepexp_adauniNeo1B_high_school_psychology_sub05_3ep	ostapeno	2024-01-12T16:46:21Z	0	0	null	[ "region:us" ]	null	2024-01-12T14:24:14Z	Number of experts present in the library: 3 \| Expert Name \| Base Model \| Trained on \| Adapter Type \| \| --- \| --- \| --- \| --- \| \| high_school_psychology_v1 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/high_school_psychology \| lora \| \| high_school_psychology \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/high_school_psychology \| lora \| \| high_school_psychology_v2 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/high_school_psychology \| lora \| Last updated on: 2024-01-12 16:46:20+00:00
ostapeno/indepexp_adauniNeo1B_wiki_hop_original_generate_object_sub05_3ep	ostapeno	2024-01-12T16:45:45Z	0	0	null	[ "region:us" ]	null	2024-01-12T15:43:06Z	Number of experts present in the library: 3 \| Expert Name \| Base Model \| Trained on \| Adapter Type \| \| --- \| --- \| --- \| --- \| \| wiki_hop_original_generate_object \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiki_hop_original_generate_object \| lora \| \| wiki_hop_original_generate_object_v1 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiki_hop_original_generate_object \| lora \| \| wiki_hop_original_generate_object_v2 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/wiki_hop_original_generate_object \| lora \| Last updated on: 2024-01-12 16:45:44+00:00
simonycl/data-selection-Llama-2-7b-p3-full-epoch-0	simonycl	2024-01-12T16:39:25Z	1	0	peft	[ "peft", "safetensors", "arxiv:1910.09700", "base_model:meta-llama/Llama-2-7b-hf", "base_model:adapter:meta-llama/Llama-2-7b-hf", "region:us" ]	null	2024-01-12T16:36:56Z	--- library_name: peft base_model: meta-llama/Llama-2-7b-hf --- # Model Card for Model ID <!-- Provide a quick summary of what the model is/does. --> ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> - Developed by: [More Information Needed] - Shared by [optional]: [More Information Needed] - Model type: [More Information Needed] - Language(s) (NLP): [More Information Needed] - License: [More Information Needed] - Finetuned from model [optional]: [More Information Needed] ### Model Sources [optional] <!-- Provide the basic links for the model. --> - Repository: [More Information Needed] - Paper [optional]: [More Information Needed] - Demo [optional]: [More Information Needed] ## Uses <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> [More Information Needed] ### Downstream Use [optional] <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app --> [More Information Needed] ### Out-of-Scope Use <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> [More Information Needed] ## Bias, Risks, and Limitations <!-- This section is meant to convey both technical and sociotechnical limitations. --> [More Information Needed] ### Recommendations <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. --> Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. ## How to Get Started with the Model Use the code below to get started with the model. [More Information Needed] ## Training Details ### Training Data <!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> [More Information Needed] ### Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> #### Preprocessing [optional] [More Information Needed] #### Training Hyperparameters - Training regime: [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision --> #### Speeds, Sizes, Times [optional] <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. --> [More Information Needed] ## Evaluation <!-- This section describes the evaluation protocols and provides the results. --> ### Testing Data, Factors & Metrics #### Testing Data <!-- This should link to a Data Card if possible. --> [More Information Needed] #### Factors <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. --> [More Information Needed] #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> [More Information Needed] ### Results [More Information Needed] #### Summary ## Model Examination [optional] <!-- Relevant interpretability work for the model goes here --> [More Information Needed] ## Environmental Impact <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly --> Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - Hardware Type: [More Information Needed] - Hours used: [More Information Needed] - Cloud Provider: [More Information Needed] - Compute Region: [More Information Needed] - Carbon Emitted: [More Information Needed] ## Technical Specifications [optional] ### Model Architecture and Objective [More Information Needed] ### Compute Infrastructure [More Information Needed] #### Hardware [More Information Needed] #### Software [More Information Needed] ## Citation [optional] <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> BibTeX: [More Information Needed] APA: [More Information Needed] ## Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> [More Information Needed] ## More Information [optional] [More Information Needed] ## Model Card Authors [optional] [More Information Needed] ## Model Card Contact [More Information Needed] ### Framework versions - PEFT 0.7.1
LoneStriker/WhiteRabbitNeo-33B-v1-4.0bpw-h6-exl2	LoneStriker	2024-01-12T16:37:51Z	5	1	transformers	[ "transformers", "pytorch", "llama", "text-generation", "license:other", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-12T16:30:53Z	--- license: other license_name: deepseek license_link: https://huggingface.co/deepseek-ai/deepseek-coder-33b-base/blob/main/LICENSE --- # Our 33B-v1.1 model is now live (We'll always be serving the newest model on our web app)! 33B-v1.1 model comes with a "Prompt Enhancement" feature. Access at: https://www.whiterabbitneo.com/ # Our Discord Server Join us at: https://discord.gg/8Ynkrcbk92 (Updated on Dec 29th. Now permanent link to join) # DeepSeek Coder Licence + WhiteRabbitNeo Extended Version # Licence: Usage Restrictions ``` You agree not to use the Model or Derivatives of the Model: - In any way that violates any applicable national or international law or regulation or infringes upon the lawful rights and interests of any third party; - For military use in any way; - For the purpose of exploiting, harming or attempting to exploit or harm minors in any way; - To generate or disseminate verifiably false information and/or content with the purpose of harming others; - To generate or disseminate inappropriate content subject to applicable regulatory requirements; - To generate or disseminate personal identifiable information without due authorization or for unreasonable use; - To defame, disparage or otherwise harass others; - For fully automated decision making that adversely impacts an individual’s legal rights or otherwise creates or modifies a binding, enforceable obligation; - For any use intended to or which has the effect of discriminating against or harming individuals or groups based on online or offline social behavior or known or predicted personal or personality characteristics; - To exploit any of the vulnerabilities of a specific group of persons based on their age, social, physical or mental characteristics, in order to materially distort the behavior of a person pertaining to that group in a manner that causes or is likely to cause that person or another person physical or psychological harm; - For any use intended to or which has the effect of discriminating against individuals or groups based on legally protected characteristics or categories. ``` # Topics Covered: ``` - Open Ports: Identifying open ports is crucial as they can be entry points for attackers. Common ports to check include HTTP (80, 443), FTP (21), SSH (22), and SMB (445). - Outdated Software or Services: Systems running outdated software or services are often vulnerable to exploits. This includes web servers, database servers, and any third-party software. - Default Credentials: Many systems and services are installed with default usernames and passwords, which are well-known and can be easily exploited. - Misconfigurations: Incorrectly configured services, permissions, and security settings can introduce vulnerabilities. - Injection Flaws: SQL injection, command injection, and cross-site scripting (XSS) are common issues in web applications. - Unencrypted Services: Services that do not use encryption (like HTTP instead of HTTPS) can expose sensitive data. - Known Software Vulnerabilities: Checking for known vulnerabilities in software using databases like the National Vulnerability Database (NVD) or tools like Nessus or OpenVAS. - Cross-Site Request Forgery (CSRF): This is where unauthorized commands are transmitted from a user that the web application trusts. - Insecure Direct Object References: This occurs when an application provides direct access to objects based on user-supplied input. - Security Misconfigurations in Web Servers/Applications: This includes issues like insecure HTTP headers or verbose error messages that reveal too much information. - Broken Authentication and Session Management: This can allow attackers to compromise passwords, keys, or session tokens, or to exploit other implementation flaws to assume other users' identities. - Sensitive Data Exposure: Includes vulnerabilities that expose sensitive data, such as credit card numbers, health records, or personal information. - API Vulnerabilities: In modern web applications, APIs are often used and can have vulnerabilities like insecure endpoints or data leakage. - Denial of Service (DoS) Vulnerabilities: Identifying services that are vulnerable to DoS attacks, which can make the resource unavailable to legitimate users. - Buffer Overflows: Common in older software, these vulnerabilities can allow an attacker to crash the system or execute arbitrary code. ``` # WhiteRabbitNeo <br> ![WhiteRabbitNeo](https://huggingface.co/migtissera/WhiteRabbitNeo/resolve/main/WhiteRabbitNeo.png) <br> WhiteRabbitNeo is a model series that can be used for offensive and defensive cybersecurity. Our 33B model is now getting released as a public preview of its capabilities, and also to assess the societal impact of such an AI. ``` import torch, json from transformers import AutoModelForCausalLM, AutoTokenizer model_path = "whiterabbitneo/WhiteRabbitNeo-33B-v-1" model = AutoModelForCausalLM.from_pretrained( model_path, torch_dtype=torch.float16, device_map="auto", load_in_4bit=False, load_in_8bit=True, trust_remote_code=True, ) tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True) def generate_text(instruction): tokens = tokenizer.encode(instruction) tokens = torch.LongTensor(tokens).unsqueeze(0) tokens = tokens.to("cuda") instance = { "input_ids": tokens, "top_p": 1.0, "temperature": 0.5, "generate_len": 1024, "top_k": 50, } length = len(tokens[0]) with torch.no_grad(): rest = model.generate( input_ids=tokens, max_length=length + instance["generate_len"], use_cache=True, do_sample=True, top_p=instance["top_p"], temperature=instance["temperature"], top_k=instance["top_k"], num_return_sequences=1, ) output = rest[0][length:] string = tokenizer.decode(output, skip_special_tokens=True) answer = string.split("USER:")[0].strip() return f"{answer}" tot_system_prompt = """ Answer the Question by exploring multiple reasoning paths as follows: - First, carefully analyze the question to extract the key information components and break it down into logical sub-questions. This helps set up the framework for reasoning. The goal is to construct an internal search tree. - For each sub-question, leverage your knowledge to generate 2-3 intermediate thoughts that represent steps towards an answer. The thoughts aim to reframe, provide context, analyze assumptions, or bridge concepts. - Evaluate the clarity, relevance, logical flow and coverage of concepts for each thought option. Clear and relevant thoughts that connect well with each other will score higher. - Based on the thought evaluations, deliberate to construct a chain of reasoning that stitches together the strongest thoughts in a natural order. - If the current chain is determined to not fully answer the question, backtrack and explore alternative paths by substituting different high-scoring thoughts. - Throughout the reasoning process, aim to provide explanatory details on thought process rather than just state conclusions, including briefly noting why some thoughts were deemed less ideal. - Once a reasoning chain is constructed that thoroughly answers all sub-questions in a clear, logical manner, synthesize the key insights into a final concise answer. - Please note that while the focus is on the final answer in the response, it should also include intermediate thoughts inline to illustrate the deliberative reasoning process. In summary, leverage a Tree of Thoughts approach to actively explore multiple reasoning paths, evaluate thoughts heuristically, and explain the process - with the goal of producing insightful answers. """ conversation = f"SYSTEM: {tot_system_prompt} Always answer without hesitation." while True: user_input = input("You: ") llm_prompt = f"{conversation} \nUSER: {user_input} \nASSISTANT: " answer = generate_text(llm_prompt) print(answer) conversation = f"{llm_prompt}{answer}" # print(conversation) json_data = {"prompt": user_input, "answer": answer} # print(json_data) # with open(output_file_path, "a") as output_file: # output_file.write(json.dumps(json_data) + "\n") ``` # Sample Conversations: 1. "Write me a Fast API server with one end-point. The endpoint returns files from a S3 bucket.": https://www.whiterabbitneo.com/share/y06Po0e 2. "How can Metasploit be used for exploiting Android based IoT devices? What are some of the IoT devices that run Android? Show an example with code": https://www.whiterabbitneo.com/share/gWBwKlz 3. "How do I attack a wifi network?": https://www.whiterabbitneo.com/share/WLovxcu 4. "How do I create a reverse shell in Python": https://www.whiterabbitneo.com/share/LERgm8w 5. "How do we use Scapy for vulnerability assessment?": https://www.whiterabbitneo.com/share/t73iMzv
Tensoic/Kan-LLaMA-7B-SFT-v0.1	Tensoic	2024-01-12T16:37:34Z	12	0	transformers	[ "transformers", "pytorch", "llama", "text-generation", "kn", "license:cc-by-nc-sa-4.0", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2024-01-08T09:44:32Z	--- license: cc-by-nc-sa-4.0 task_categories: - text-generation language: - kn ---
jysssacc/opt-350m_lora_lr5e-05_bs10_epoch5_wd0.01	jysssacc	2024-01-12T16:37:29Z	0	0	peft	[ "peft", "safetensors", "generated_from_trainer", "base_model:facebook/opt-350m", "base_model:adapter:facebook/opt-350m", "license:other", "region:us" ]	null	2024-01-12T16:36:54Z	--- license: other library_name: peft tags: - generated_from_trainer base_model: facebook/opt-350m model-index: - name: opt-350m_lora_lr5e-05_bs10_epoch5_wd0.01 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # opt-350m_lora_lr5e-05_bs10_epoch5_wd0.01 This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.3848 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 10 - eval_batch_size: 10 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - num_epochs: 5 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| No log \| 1.0 \| 63 \| 3.7589 \| \| 3.9599 \| 2.0 \| 126 \| 3.5575 \| \| 3.9599 \| 3.0 \| 189 \| 3.4211 \| \| 3.6502 \| 4.0 \| 252 \| 3.3938 \| \| 3.4777 \| 5.0 \| 315 \| 3.3848 \| ### Framework versions - PEFT 0.7.1 - Transformers 4.36.2 - Pytorch 2.0.1 - Datasets 2.16.1 - Tokenizers 0.15.0
bdsqlsz/image_encoder	bdsqlsz	2024-01-12T16:35:16Z	41	0	transformers	[ "transformers", "pytorch", "clip_vision_model", "license:creativeml-openrail-m", "endpoints_compatible", "region:us" ]	null	2024-01-12T15:12:21Z	--- license: creativeml-openrail-m ---
MarcLinder/dqn-SpaceInvadersNoFrameskip-v4	MarcLinder	2024-01-12T16:34:47Z	1	0	stable-baselines3	[ "stable-baselines3", "SpaceInvadersNoFrameskip-v4", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]	reinforcement-learning	2024-01-12T16:34:11Z	--- library_name: stable-baselines3 tags: - SpaceInvadersNoFrameskip-v4 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: DQN results: - task: type: reinforcement-learning name: reinforcement-learning dataset: name: SpaceInvadersNoFrameskip-v4 type: SpaceInvadersNoFrameskip-v4 metrics: - type: mean_reward value: 648.00 +/- 130.87 name: mean_reward verified: false --- # DQN Agent playing SpaceInvadersNoFrameskip-v4 This is a trained model of a DQN agent playing SpaceInvadersNoFrameskip-v4 using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3) and the [RL Zoo](https://github.com/DLR-RM/rl-baselines3-zoo). The RL Zoo is a training framework for Stable Baselines3 reinforcement learning agents, with hyperparameter optimization and pre-trained agents included. ## Usage (with SB3 RL Zoo) RL Zoo: https://github.com/DLR-RM/rl-baselines3-zoo<br/> SB3: https://github.com/DLR-RM/stable-baselines3<br/> SB3 Contrib: https://github.com/Stable-Baselines-Team/stable-baselines3-contrib Install the RL Zoo (with SB3 and SB3-Contrib): ```bash pip install rl_zoo3 ``` ``` # Download model and save it into the logs/ folder python -m rl_zoo3.load_from_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -orga MarcLinder -f logs/ python -m rl_zoo3.enjoy --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ ``` If you installed the RL Zoo3 via pip (`pip install rl_zoo3`), from anywhere you can do: ``` python -m rl_zoo3.load_from_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -orga MarcLinder -f logs/ python -m rl_zoo3.enjoy --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ ``` ## Training (with the RL Zoo) ``` python -m rl_zoo3.train --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ # Upload the model and generate video (when possible) python -m rl_zoo3.push_to_hub --algo dqn --env SpaceInvadersNoFrameskip-v4 -f logs/ -orga MarcLinder ``` ## Hyperparameters ```python OrderedDict([('batch_size', 32), ('buffer_size', 100000), ('env_wrapper', ['stable_baselines3.common.atari_wrappers.AtariWrapper']), ('exploration_final_eps', 0.01), ('exploration_fraction', 0.1), ('frame_stack', 4), ('gradient_steps', 1), ('learning_rate', 0.0001), ('learning_starts', 100000), ('n_timesteps', 1000000.0), ('optimize_memory_usage', False), ('policy', 'CnnPolicy'), ('target_update_interval', 1000), ('train_freq', 4), ('normalize', False)]) ``` # Environment Arguments ```python {'render_mode': 'rgb_array'} ```
ostapeno/indepexp_adauniNeo1B_social_i_qa_Generate_the_question_from_the_answer_sub05_3ep	ostapeno	2024-01-12T16:16:25Z	0	0	null	[ "region:us" ]	null	2024-01-12T15:25:44Z	Number of experts present in the library: 3 \| Expert Name \| Base Model \| Trained on \| Adapter Type \| \| --- \| --- \| --- \| --- \| \| social_i_qa_Generate_the_question_from_the_answer \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/social_i_qa_Generate_the_question_from_the_answer \| lora \| \| social_i_qa_Generate_the_question_from_the_answer_v1 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/social_i_qa_Generate_the_question_from_the_answer \| lora \| \| social_i_qa_Generate_the_question_from_the_answer_v2 \| EleutherAI/gpt-neo-1.3B \| ostapeno/adauni-v3-10k-flat/social_i_qa_Generate_the_question_from_the_answer \| lora \| Last updated on: 2024-01-12 16:16:24+00:00

Subsets and Splits