modelId
stringlengths 5
139
| author
stringlengths 2
42
| last_modified
timestamp[us, tz=UTC]date 2020-02-15 11:33:14
2025-07-14 12:27:51
| downloads
int64 0
223M
| likes
int64 0
11.7k
| library_name
stringclasses 520
values | tags
listlengths 1
4.05k
| pipeline_tag
stringclasses 55
values | createdAt
timestamp[us, tz=UTC]date 2022-03-02 23:29:04
2025-07-14 12:25:52
| card
stringlengths 11
1.01M
|
---|---|---|---|---|---|---|---|---|---|
DrishtiSharma/llama-pro-8b-tweet-summarization-gradnorm-0.3 | DrishtiSharma | 2024-02-01T08:21:24Z | 1 | 0 | peft | [
"peft",
"safetensors",
"trl",
"sft",
"generated_from_trainer",
"dataset:dialogstudio",
"base_model:TencentARC/LLaMA-Pro-8B",
"base_model:adapter:TencentARC/LLaMA-Pro-8B",
"license:llama2",
"region:us"
]
| null | 2024-02-01T06:31:57Z | ---
license: llama2
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- dialogstudio
base_model: TencentARC/LLaMA-Pro-8B
model-index:
- name: llama-pro-8b-tweet-summarization-gradnorm-0.3
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. -->
# llama-pro-8b-tweet-summarization-gradnorm-0.3
This model is a fine-tuned version of [TencentARC/LLaMA-Pro-8B](https://huggingface.co/TencentARC/LLaMA-Pro-8B) on the dialogstudio dataset.
It achieves the following results on the evaluation set:
- Loss: 2.9796
- Rouge Scores: {'rouge1': 93.71888929189157, 'rouge2': 77.8377567936117, 'rougeL': 64.47906852741538, 'rougeLsum': 93.71298018429633}
- Bleu Scores: [0.9470990193868204, 0.9341779145832757, 0.9064440397746264, 0.8744914403659334]
- Gen Len: 463.0182
## 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: cosine
- num_epochs: 7
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge Scores | Bleu Scores | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:------------------------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------:|:--------:|
| 1.9065 | 1.0 | 220 | 1.8530 | {'rouge1': 92.83694737064799, 'rouge2': 78.72458121869542, 'rougeL': 67.88788283384865, 'rougeLsum': 92.83768512059282} | [0.8739198483584956, 0.8530170264142946, 0.8271978418182495, 0.7998377773703629] | 463.0182 |
| 1.6363 | 2.0 | 440 | 1.8633 | {'rouge1': 93.54135671371444, 'rouge2': 78.96116387599493, 'rougeL': 67.77857901494997, 'rougeLsum': 93.54432289584433} | [0.8758125801988195, 0.8577741180618648, 0.8322886881519586, 0.80457236049974] | 463.0182 |
| 1.2817 | 3.0 | 660 | 2.0098 | {'rouge1': 87.30764070509844, 'rouge2': 73.12328274037898, 'rougeL': 62.00625532521349, 'rougeLsum': 87.29149649901954} | [0.8757949025917542, 0.8593181834244542, 0.8334473061685955, 0.8048319452251607] | 463.0182 |
| 0.9049 | 4.0 | 880 | 2.2481 | {'rouge1': 87.35996946418575, 'rouge2': 72.87802745947901, 'rougeL': 61.35206821444361, 'rougeLsum': 87.32662841081371} | [0.8755472589597261, 0.859572654041077, 0.8333237300074641, 0.804082483213136] | 463.0182 |
| 0.5916 | 5.0 | 1100 | 2.5061 | {'rouge1': 78.38431994557745, 'rouge2': 64.89809559762811, 'rougeL': 53.805209482421525, 'rougeLsum': 78.30608179426231} | [0.747179346815877, 0.7352208249958618, 0.7126103103040894, 0.6869428956670465] | 463.0182 |
| 0.3898 | 6.0 | 1320 | 2.8150 | {'rouge1': 93.77539618029996, 'rouge2': 78.03050568501187, 'rougeL': 64.82344374456906, 'rougeLsum': 93.76894400818286} | [0.9469183628254614, 0.9342162110956728, 0.9067374010427977, 0.8750430150656403] | 463.0182 |
| 0.2961 | 7.0 | 1540 | 2.9796 | {'rouge1': 93.71888929189157, 'rouge2': 77.8377567936117, 'rougeL': 64.47906852741538, 'rougeLsum': 93.71298018429633} | [0.9470990193868204, 0.9341779145832757, 0.9064440397746264, 0.8744914403659334] | 463.0182 |
### Framework versions
- PEFT 0.8.2.dev0
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.2.dev0
- Tokenizers 0.15.1 |
vitruv/vitruv_1 | vitruv | 2024-02-01T08:19:08Z | 116 | 0 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"ko",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T07:48:27Z | ---
license: apache-2.0
language:
- ko
---
Who we are : Virtruv
ํด๋น ๋ชจ๋ธ์ ํ๊ตญ์ด ์ค ์ํ ๋ชจ๋ธ์ ์ง์คํ์ฌ ํ์ต์ ์๋ํ์์ต๋๋ค.
Base Model : 'beomi/OPEN-SOLAR-KO-10.7B'
Dataset :
1 . traintogpb/aihub-koen-translation-integrated-tiny-100k
2. kyujinpy/KOR-gugugu-platypus-set
3. GAIR/MathPile : ๋ค์ ๋ฐ์ดํฐ ์
์ sampling ํ์ฌ ์ง์ translate, ํ์์ต๋๋ค.
Prompt:
|
DrishtiSharma/llama2-7bb-tweet-summarization-gradnorm-0.3 | DrishtiSharma | 2024-02-01T08:17:42Z | 0 | 0 | peft | [
"peft",
"safetensors",
"trl",
"sft",
"generated_from_trainer",
"dataset:dialogstudio",
"base_model:NousResearch/Llama-2-7b-hf",
"base_model:adapter:NousResearch/Llama-2-7b-hf",
"region:us"
]
| null | 2024-02-01T08:16:54Z | ---
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- dialogstudio
base_model: NousResearch/Llama-2-7b-hf
model-index:
- name: llama2-7bb-tweet-summarization-gradnorm-0.3
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. -->
# llama2-7bb-tweet-summarization-gradnorm-0.3
This model is a fine-tuned version of [NousResearch/Llama-2-7b-hf](https://huggingface.co/NousResearch/Llama-2-7b-hf) on the dialogstudio dataset.
It achieves the following results on the evaluation set:
- Loss: 2.8160
- Rouge Scores: {'rouge1': 93.719779910895, 'rouge2': 78.0799701185797, 'rougeL': 64.91384075272471, 'rougeLsum': 93.71249369436103}
- Bleu Scores: [0.9468715981421053, 0.9340571158071639, 0.906767913949756, 0.8753561378232885]
- Gen Len: 463.0182
## 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: cosine
- num_epochs: 7
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge Scores | Bleu Scores | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:-----------------------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------:|:--------:|
| 1.9246 | 1.0 | 220 | 1.8384 | {'rouge1': 92.78080137059182, 'rouge2': 78.71532643138437, 'rougeL': 68.0616149947273, 'rougeLsum': 92.78702835703021} | [0.9079275318266272, 0.8970741286020552, 0.8736002135507472, 0.8466150307526832] | 463.0182 |
| 1.6564 | 2.0 | 440 | 1.8335 | {'rouge1': 93.62527163754612, 'rouge2': 79.14899366889107, 'rougeL': 68.02122989340602, 'rougeLsum': 93.62676386700348} | [0.9282164809556785, 0.9171615801879893, 0.892709310950969, 0.8645188775345913] | 463.0182 |
| 1.3403 | 3.0 | 660 | 1.9481 | {'rouge1': 93.70688850262614, 'rouge2': 78.96026100012381, 'rougeL': 67.37638965440908, 'rougeLsum': 93.70399692691778} | [0.9342903619020663, 0.9225682522334384, 0.8972845918789121, 0.8681853449069523] | 463.0182 |
| 0.9984 | 4.0 | 880 | 2.1537 | {'rouge1': 93.77800041953847, 'rouge2': 78.72204799373465, 'rougeL': 66.56763131340682, 'rougeLsum': 93.77100407824561} | [0.9425931953005738, 0.9302863494509406, 0.9040669212466305, 0.8739193334758137] | 463.0182 |
| 0.7 | 5.0 | 1100 | 2.3692 | {'rouge1': 93.74639046979189, 'rouge2': 78.51569240275262, 'rougeL': 65.93032986525995, 'rougeLsum': 93.73745084400457} | [0.9440175755443134, 0.93171453625075, 0.9052208696375351, 0.8747208115562404] | 463.0182 |
| 0.4947 | 6.0 | 1320 | 2.6590 | {'rouge1': 93.75661844384149, 'rouge2': 78.18805763398609, 'rougeL': 65.29243896759789, 'rougeLsum': 93.75034348574664} | [0.9470358425741272, 0.9342995624545122, 0.9070823690393129, 0.8757451333358709] | 463.0182 |
| 0.3922 | 7.0 | 1540 | 2.8160 | {'rouge1': 93.719779910895, 'rouge2': 78.0799701185797, 'rougeL': 64.91384075272471, 'rougeLsum': 93.71249369436103} | [0.9468715981421053, 0.9340571158071639, 0.906767913949756, 0.8753561378232885] | 463.0182 |
### Framework versions
- PEFT 0.8.2.dev0
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.2.dev0
- Tokenizers 0.15.1 |
Sacralet/dbw-bert-large-1.1 | Sacralet | 2024-02-01T08:16:00Z | 2 | 0 | transformers | [
"transformers",
"safetensors",
"bert",
"fill-mask",
"generated_from_trainer",
"base_model:google-bert/bert-large-uncased",
"base_model:finetune:google-bert/bert-large-uncased",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| fill-mask | 2024-02-01T02:57:28Z | ---
license: apache-2.0
base_model: bert-large-uncased
tags:
- generated_from_trainer
model-index:
- name: dbw-bert-large-1.1
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. -->
# dbw-bert-large-1.1
This model is a fine-tuned version of [bert-large-uncased](https://huggingface.co/bert-large-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0263
## 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: 32
- eval_batch_size: 8
- seed: 42
- 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 |
|:-------------:|:-----:|:----:|:---------------:|
| 0.0616 | 0.09 | 100 | 0.0571 |
| 0.0516 | 0.18 | 200 | 0.0438 |
| 0.0417 | 0.27 | 300 | 0.0349 |
| 0.0372 | 0.36 | 400 | 0.0319 |
| 0.0317 | 0.44 | 500 | 0.0306 |
| 0.0309 | 0.53 | 600 | 0.0294 |
| 0.0287 | 0.62 | 700 | 0.0286 |
| 0.029 | 0.71 | 800 | 0.0273 |
| 0.0351 | 0.8 | 900 | 0.0265 |
| 0.0252 | 0.89 | 1000 | 0.0262 |
| 0.0226 | 0.98 | 1100 | 0.0263 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
cecb/newsfinetune | cecb | 2024-02-01T08:14:58Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| null | 2024-02-01T08:14:46Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
agshubhi/Insurance_complaint_mgmt | agshubhi | 2024-02-01T08:13:12Z | 0 | 0 | null | [
"dataset:ebrigham/NL_insurance_reviews_sentiment",
"license:mit",
"region:us"
]
| null | 2024-02-01T07:51:27Z | ---
license: mit
datasets:
- ebrigham/NL_insurance_reviews_sentiment
--- |
Shreyas0706/Llama-2-7b-chat-hf-fine-tuned-adapters | Shreyas0706 | 2024-02-01T07:59:43Z | 0 | 0 | peft | [
"peft",
"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 | 2024-02-01T07:59:37Z | ---
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 0.8.2.dev0 |
LoneStriker/CapybaraHermes-2.5-Mistral-7B-3.0bpw-h6-exl2 | LoneStriker | 2024-02-01T07:58:06Z | 8 | 0 | trl | [
"trl",
"safetensors",
"mistral",
"distilabel",
"dpo",
"rlaif",
"rlhf",
"en",
"dataset:argilla/dpo-mix-7k",
"license:apache-2.0",
"region:us"
]
| null | 2024-02-01T07:56:25Z | ---
library_name: trl
license: apache-2.0
datasets:
- argilla/dpo-mix-7k
language:
- en
tags:
- distilabel
- dpo
- rlaif
- rlhf
---
# CapybaraHermes-2.5-Mistral-7B
<div>
<img src="https://cdn-uploads.huggingface.co/production/uploads/60420dccc15e823a685f2b03/Vmr0FtTvnny6Snm-UDM_n.png">
</div>
<p align="center">
<a href="https://github.com/argilla-io/distilabel">
<img src="https://raw.githubusercontent.com/argilla-io/distilabel/main/docs/assets/distilabel-badge-light.png" alt="Built with Distilabel" width="200" height="32"/>
</a>
</p>
This model is the launching partner of the [capybara-dpo dataset](https://huggingface.co/datasets/argilla/distilabel-capybara-dpo-9k-binarized) build with โ๏ธ distilabel. It's a preference tuned [OpenHermes-2.5-Mistral-7B](https://huggingface.co/teknium/OpenHermes-2.5-Mistral-7B).
CapybaraHermes has been preference tuned with LoRA and TRL for 3 epochs using argilla's [dpo mix 7k](https://huggingface.co/datasets/argilla/dpo-mix-7k).
To test the impact on multi-turn performance we have used MTBench. We also include the Nous Benchmark results and Mistral-7B-Instruct-v0.2 for reference as it's a strong 7B model on MTBench:
| Model | AGIEval | GPT4All | TruthfulQA | Bigbench | MTBench First Turn | MTBench Second Turn | Nous avg. | MTBench avg. |
|-----------------------------------|---------|---------|------------|----------|------------|-------------|-----------|--------------|
| argilla/CapybaraHermes-2.5-Mistral-7B | **43.8** | **73.35** | 57.07 | **42.44** | 8.24375 | **7.5625** | 54.16 | **7.903125** |
| teknium/OpenHermes-2.5-Mistral-7B | 42.75 | 72.99 | 52.99 | 40.94 | **8.25** | 7.2875 | 52.42 | 7.76875 |
| Mistral-7B-Instruct-v0.2 | 38.5 | 71.64 | **66.82** | 42.29 | 7.8375 | 7.1 | **54.81** | 7.46875 |
The most interesting aspect in the context of the capybara-dpo dataset is the increased performance in MTBench Second Turn scores.
For the merge lovers, we also preference tuned Beagle14-7B with a mix of capybara-dpo and distilabel orca pairs using the same recipe as NeuralBeagle (see [ YALL - Yet Another LLM Leaderboard](https://huggingface.co/spaces/mlabonne/Yet_Another_LLM_Leaderboard) for reference):
| Model |AGIEval|GPT4All|TruthfulQA|Bigbench|Average|
|------------------------------------------------------------------------------------------------------------------------------------|------:|------:|---------:|-------:|------:|
|[DistilabelBeagle14-7B](https://huggingface.co/dvilasuero/DistilabelBeagle14-7B)| 45.29| 76.92| 71.66| 48.78| 60.66|
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** Argilla
- **Shared by [optional]:** Argilla
- **Model type:** 7B chat model
- **Language(s) (NLP):** English
- **License:** Same as OpenHermes
- **Finetuned from model [optional]:** [OpenHermes-2.5-Mistral-7B](https://huggingface.co/teknium/OpenHermes-2.5-Mistral-7B)
|
franklee1015/ppo-LunarLander-v2 | franklee1015 | 2024-02-01T07:58:04Z | 1 | 0 | stable-baselines3 | [
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-01-29T03:40:37Z | ---
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: 298.43 +/- 12.85
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
...
```
|
mkdir700/v3-starcoderbase1b-personal-copilot-A100-40GB-colab | mkdir700 | 2024-02-01T07:53:07Z | 184 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"gpt_bigcode",
"text-generation",
"generated_from_trainer",
"base_model:bigcode/starcoderbase-1b",
"base_model:finetune:bigcode/starcoderbase-1b",
"license:bigcode-openrail-m",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T06:24:51Z | ---
license: bigcode-openrail-m
tags:
- generated_from_trainer
base_model: bigcode/starcoderbase-1b
model-index:
- name: v3-starcoderbase1b-personal-copilot-A100-40GB-colab
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. -->
# v3-starcoderbase1b-personal-copilot-A100-40GB-colab
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: nan
## 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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- 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 |
|:-------------:|:-----:|:----:|:---------------:|
| 0.0 | 0.05 | 100 | nan |
| 0.0 | 0.1 | 200 | nan |
| 0.0 | 0.15 | 300 | nan |
| 0.0 | 0.2 | 400 | nan |
| 0.0 | 0.25 | 500 | nan |
| 0.0 | 0.3 | 600 | nan |
| 0.0 | 0.35 | 700 | nan |
| 0.0 | 0.4 | 800 | nan |
| 0.0 | 0.45 | 900 | nan |
| 0.0 | 0.5 | 1000 | nan |
| 0.0 | 0.55 | 1100 | nan |
| 0.0 | 0.6 | 1200 | nan |
| 0.0 | 0.65 | 1300 | nan |
| 0.0 | 0.7 | 1400 | nan |
| 0.0 | 0.75 | 1500 | nan |
| 0.0 | 0.8 | 1600 | nan |
| 0.0 | 0.85 | 1700 | nan |
| 0.0 | 0.9 | 1800 | nan |
| 0.0 | 0.95 | 1900 | nan |
| 0.0 | 1.0 | 2000 | nan |
### Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
dhanikitkat/topic-class-indo-bank | dhanikitkat | 2024-02-01T07:43:28Z | 126 | 0 | transformers | [
"transformers",
"pytorch",
"tf",
"safetensors",
"roberta",
"text-classification",
"indonesian-roberta-base-sentiment-classifier",
"id",
"dataset:indonlu",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-classification | 2024-01-31T04:31:49Z | ---
language: id
tags:
- indonesian-roberta-base-sentiment-classifier
license: mit
datasets:
- indonlu
widget:
- text: "di update malah tidak bisa dibuka"
--- |
THUDM/LongAlign-13B-64k-base | THUDM | 2024-02-01T07:31:38Z | 21 | 3 | transformers | [
"transformers",
"pytorch",
"llama",
"text-generation",
"Long Context",
"en",
"zh",
"arxiv:2401.18058",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-29T13:37:47Z | ---
language:
- en
- zh
library_name: transformers
tags:
- Long Context
- llama
pipeline_tag: text-generation
license: apache-2.0
---
# LongAlign-13B-64k-base
<p align="center">
๐ค <a href="https://huggingface.co/datasets/THUDM/LongAlign-10k" target="_blank">[LongAlign Dataset] </a> โข ๐ป <a href="https://github.com/THUDM/LongAlign" target="_blank">[Github Repo]</a> โข ๐ <a href="https://arxiv.org/abs/2401.18058" target="_blank">[LongAlign Paper]</a>
</p>
**LongAlign** is the first full recipe for LLM alignment on long context. We propose the **LongAlign-10k** dataset, containing 10,000 long instruction data of 8k-64k in length. We investigate on trianing strategies, namely **packing (with loss weighting) and sorted batching**, which are all implemented in our code. For real-world long context evaluation, we introduce **LongBench-Chat** that evaluate the instruction-following capability on queries of 10k-100k length.
## All Models
We open-sourced the following list of models:
|Model|Huggingface Repo|Description|
|---|---|---|
|**LongAlign-6B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k-base) | **ChatGLM3-6B** with an extended 64k context window |
|**LongAlign-6B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k) | Chat model by LongAlign training on LongAlign-6B-64k-base|
|**LongAlign-7B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k-base) | **Llama-2-7B** with an extended 64k context window |
|**LongAlign-7B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k) | Chat model by LongAlign training on LongAlign-7B-64k-base|
|**LongAlign-13B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k-base) | **Llama-2-13B** with an extended 64k context window |
|**LongAlign-13B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k) | Chat model by LongAlign training on LongAlign-13B-64k-base|
|**ChatGLM3-6B-128k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/chatglm3-6b-128k) | **ChatGLM3-6B** with a 128k context window|

## Model usage
Chat prompt template for LongAlign-6B-64k:
```text
[Round 1]
้ฎ๏ผHi!
็ญ๏ผHello! What can I assist you today?
[Round 2]
้ฎ๏ผWhat should I do if I can't sleep at night?
็ญ๏ผ
```
Chat prompt template for LongAlign-7B-64k and LongAlign-13B-64k:
```text
[INST]Hi![/INST]Hello! What can I assist you today?
[INST]What should I do if I can't sleep at night?[/INST]
```
ChatGLM3-6B-128k uses the same prompt template as [ChatGLM3-6B](https://huggingface.co/THUDM/chatglm3-6b).
A simple demo for deployment of the model:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
tokenizer = AutoTokenizer.from_pretrained("THUDM/LongAlign-6B-64k", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("THUDM/LongAlign-6B-64k", torch_dtype=torch.bfloat16, trust_remote_code=True, device_map="auto")
model = model.eval()
query = open("assets/paper.txt").read() + "\n\nPlease summarize the paper."
response, history = model.chat(tokenizer, query, history=[], max_new_tokens=512, temperature=1)
print(response)
```
## Citation
If you find our work useful, please consider citing LongAlign:
```
``` |
THUDM/LongAlign-13B-64k | THUDM | 2024-02-01T07:31:22Z | 22 | 13 | transformers | [
"transformers",
"pytorch",
"llama",
"text-generation",
"Long Context",
"en",
"zh",
"dataset:THUDM/LongAlign-10k",
"arxiv:2401.18058",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-29T11:29:19Z | ---
language:
- en
- zh
library_name: transformers
tags:
- Long Context
- llama
datasets:
- THUDM/LongAlign-10k
pipeline_tag: text-generation
license: apache-2.0
---
# LongAlign-13B-64k
<p align="center">
๐ค <a href="https://huggingface.co/datasets/THUDM/LongAlign-10k" target="_blank">[LongAlign Dataset] </a> โข ๐ป <a href="https://github.com/THUDM/LongAlign" target="_blank">[Github Repo]</a> โข ๐ <a href="https://arxiv.org/abs/2401.18058" target="_blank">[LongAlign Paper]</a>
</p>
**LongAlign** is the first full recipe for LLM alignment on long context. We propose the **LongAlign-10k** dataset, containing 10,000 long instruction data of 8k-64k in length. We investigate on trianing strategies, namely **packing (with loss weighting) and sorted batching**, which are all implemented in our code. For real-world long context evaluation, we introduce **LongBench-Chat** that evaluate the instruction-following capability on queries of 10k-100k length.
## All Models
We open-sourced the following list of models:
|Model|Huggingface Repo|Description|
|---|---|---|
|**LongAlign-6B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k-base) | **ChatGLM3-6B** with an extended 64k context window |
|**LongAlign-6B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k) | Chat model by LongAlign training on LongAlign-6B-64k-base|
|**LongAlign-7B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k-base) | **Llama-2-7B** with an extended 64k context window |
|**LongAlign-7B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k) | Chat model by LongAlign training on LongAlign-7B-64k-base|
|**LongAlign-13B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k-base) | **Llama-2-13B** with an extended 64k context window |
|**LongAlign-13B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k) | Chat model by LongAlign training on LongAlign-13B-64k-base|
|**ChatGLM3-6B-128k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/chatglm3-6b-128k) | **ChatGLM3-6B** with a 128k context window|

## Model usage
Chat prompt template for LongAlign-6B-64k:
```text
[Round 1]
้ฎ๏ผHi!
็ญ๏ผHello! What can I assist you today?
[Round 2]
้ฎ๏ผWhat should I do if I can't sleep at night?
็ญ๏ผ
```
Chat prompt template for LongAlign-7B-64k and LongAlign-13B-64k:
```text
[INST]Hi![/INST]Hello! What can I assist you today?
[INST]What should I do if I can't sleep at night?[/INST]
```
ChatGLM3-6B-128k uses the same prompt template as [ChatGLM3-6B](https://huggingface.co/THUDM/chatglm3-6b).
A simple demo for deployment of the model:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
tokenizer = AutoTokenizer.from_pretrained("THUDM/LongAlign-6B-64k", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("THUDM/LongAlign-6B-64k", torch_dtype=torch.bfloat16, trust_remote_code=True, device_map="auto")
model = model.eval()
query = open("assets/paper.txt").read() + "\n\nPlease summarize the paper."
response, history = model.chat(tokenizer, query, history=[], max_new_tokens=512, temperature=1)
print(response)
```
## Citation
If you find our work useful, please consider citing LongAlign:
```
``` |
THUDM/LongAlign-7B-64k | THUDM | 2024-02-01T07:30:44Z | 191 | 3 | transformers | [
"transformers",
"pytorch",
"llama",
"text-generation",
"Long Context",
"en",
"zh",
"dataset:THUDM/LongAlign-10k",
"arxiv:2401.18058",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-29T09:06:16Z | ---
language:
- en
- zh
library_name: transformers
tags:
- Long Context
- llama
datasets:
- THUDM/LongAlign-10k
license: apache-2.0
---
# LongAlign-7B-64k
<p align="center">
๐ค <a href="https://huggingface.co/datasets/THUDM/LongAlign-10k" target="_blank">[LongAlign Dataset] </a> โข ๐ป <a href="https://github.com/THUDM/LongAlign" target="_blank">[Github Repo]</a> โข ๐ <a href="https://arxiv.org/abs/2401.18058" target="_blank">[LongAlign Paper]</a>
</p>
**LongAlign** is the first full recipe for LLM alignment on long context. We propose the **LongAlign-10k** dataset, containing 10,000 long instruction data of 8k-64k in length. We investigate on trianing strategies, namely **packing (with loss weighting) and sorted batching**, which are all implemented in our code. For real-world long context evaluation, we introduce **LongBench-Chat** that evaluate the instruction-following capability on queries of 10k-100k length.
## All Models
We open-sourced the following list of models:
|Model|Huggingface Repo|Description|
|---|---|---|
|**LongAlign-6B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k-base) | **ChatGLM3-6B** with an extended 64k context window |
|**LongAlign-6B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k) | Chat model by LongAlign training on LongAlign-6B-64k-base|
|**LongAlign-7B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k-base) | **Llama-2-7B** with an extended 64k context window |
|**LongAlign-7B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k) | Chat model by LongAlign training on LongAlign-7B-64k-base|
|**LongAlign-13B-64k-base**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k-base) | **Llama-2-13B** with an extended 64k context window |
|**LongAlign-13B-64k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k) | Chat model by LongAlign training on LongAlign-13B-64k-base|
|**ChatGLM3-6B-128k**| [๐ค Huggingface Repo](https://huggingface.co/THUDM/chatglm3-6b-128k) | **ChatGLM3-6B** with a 128k context window|

## Model usage
Chat prompt template for LongAlign-6B-64k:
```text
[Round 1]
้ฎ๏ผHi!
็ญ๏ผHello! What can I assist you today?
[Round 2]
้ฎ๏ผWhat should I do if I can't sleep at night?
็ญ๏ผ
```
Chat prompt template for LongAlign-7B-64k and LongAlign-13B-64k:
```text
[INST]Hi![/INST]Hello! What can I assist you today?
[INST]What should I do if I can't sleep at night?[/INST]
```
ChatGLM3-6B-128k uses the same prompt template as [ChatGLM3-6B](https://huggingface.co/THUDM/chatglm3-6b).
A simple demo for deployment of the model:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
tokenizer = AutoTokenizer.from_pretrained("THUDM/LongAlign-6B-64k", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("THUDM/LongAlign-6B-64k", torch_dtype=torch.bfloat16, trust_remote_code=True, device_map="auto")
model = model.eval()
query = open("assets/paper.txt").read() + "\n\nPlease summarize the paper."
response, history = model.chat(tokenizer, query, history=[], max_new_tokens=512, temperature=1)
print(response)
```
## Citation
If you find our work useful, please consider citing LongAlign:
```
``` |
Sosnitskij/fialka-13B-v4-gguf | Sosnitskij | 2024-02-01T07:29:25Z | 16 | 1 | null | [
"gguf",
"text-generation",
"ru",
"en",
"dataset:0x7194633/fialka-v3-data",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-30T23:00:13Z | ---
license: apache-2.0
datasets:
- 0x7194633/fialka-v3-data
language:
- ru
- en
pipeline_tag: text-generation
---
Original model: https://huggingface.co/0x7194633/fialka-13B-v4
# Fialka v4.0 13B

## Description
Fialka language models are trained to follow instructions and maintain communication in Russian. The fourth version of Fialka is the third version optimized through RLHF. More responsive and more informative.
## Usage
The model has a query format as in zephyr.
```
<|user|>
ะงัะพ ัะฐะบะพะต ะผะตะผ?</s>
<|assistant|>
ะะตะผ (ะฐะฝะณะป. meme) - ััะพ ะตะดะธะฝะธัะฐ ัะพัะธะฐะปัะฝะพะณะพ ะฟะพะฒะตะดะตะฝะธั, ะบะพัะพัะฐั ะฑััััะพ ัะฐัะฟัะพัััะฐะฝัะตััั ะฒ ะธะฝัะตัะฝะตัะต ะธะปะธ ะฒ ัะพัะธะฐะปัะฝัั
ัะตััั
ั ัะตะปัั ะฟะตัะตะดะฐัะธ ะธะฝัะพัะผะฐัะธะธ ะธ ะธะดะตะน. ะะฑััะฝะพ ะผะตะผั ัะฒะปััััั ัััะบะฐะผะธ, ััะธัะบะฐะผะธ, ะธะทะพะฑัะฐะถะตะฝะธัะผะธ ะธะปะธ ะฒะธะดะตะพ ะธ ะธะผะตัั ัะผะพัะธััะธัะตัะบะธะน ะธะปะธ ัะฐัะธัะธัะตัะบะธะน ั
ะฐัะฐะบัะตั, ะฝะพ ะผะพะณัั ัะพะดะตัะถะฐัั ะธ ะฑะพะปะตะต ัะตััะตะทะฝัะต ะธะดะตะธ, ัะฐะบะธะต ะบะฐะบ ะฟะพะปะธัะธัะตัะบะธะต ะธะปะธ ัะพัะธะฐะปัะฝัะต ะฟัะพัะตััั, ะธ ะดะฐะถะต ัะณัะพะทั. ะะตะผั ะผะพะณัั ัะปัะถะธัั ะดะปั ัะพะทะดะฐะฝะธั ะธ ัะฐัะฟัะพัััะฐะฝะตะฝะธั ะบะพะฝัะตะฝัะฐ ะธ ะธะฝัะพัะผะฐัะธะธ, ะฐ ัะฐะบะถะต ะดะปั ะฒััะฐะถะตะฝะธั ะผะฝะตะฝะธั ะธะปะธ ััะฒััะฒ ะฐะฒัะพัะฐ.
``` |
LoneStriker/DistilabelBeagle14-7B-5.0bpw-h6-exl2 | LoneStriker | 2024-02-01T07:28:19Z | 16 | 1 | transformers | [
"transformers",
"safetensors",
"mistral",
"text-generation",
"merge",
"mergekit",
"lazymergekit",
"dpo",
"rlhf",
"rlaif",
"distilabel",
"conversational",
"arxiv:1910.09700",
"base_model:mlabonne/Beagle14-7B",
"base_model:finetune:mlabonne/Beagle14-7B",
"license:cc-by-nc-4.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T07:26:19Z | ---
license: cc-by-nc-4.0
base_model: mlabonne/Beagle14-7B
tags:
- merge
- mergekit
- lazymergekit
- dpo
- rlhf
- rlaif
- distilabel
---
# Model Card for Model ID
This is a preference tuned version of `mlabonne/Beagle14-7B` using a mix of Argilla's orca pairs and a new upcoming multi-turn dpo dataset.
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** Argilla
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** English
- **License:** cc-by-nc-4.0
- **Finetuned from model [optional]:** mlabonne/Beagle14-7B
### 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]
|
bachbouch/w-3-qlora-full | bachbouch | 2024-02-01T07:27:31Z | 5 | 0 | transformers | [
"transformers",
"safetensors",
"mistral",
"text-generation",
"conversational",
"arxiv:1910.09700",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T06:58:41Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
medinamanuel/q-FrozenLake-v1-4x4-noSlippery | medinamanuel | 2024-02-01T07:17:06Z | 0 | 0 | null | [
"FrozenLake-v1-4x4-no_slippery",
"q-learning",
"reinforcement-learning",
"custom-implementation",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T07:17:01Z | ---
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="medinamanuel/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"])
```
|
Grekkla/HubbleSpacePhotograph | Grekkla | 2024-02-01T07:13:17Z | 1 | 0 | diffusers | [
"diffusers",
"text-to-image",
"stable-diffusion",
"lora",
"template:sd-lora",
"base_model:stabilityai/stable-diffusion-xl-base-1.0",
"base_model:adapter:stabilityai/stable-diffusion-xl-base-1.0",
"license:unknown",
"region:us"
]
| text-to-image | 2024-02-01T07:11:53Z | ---
tags:
- text-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
widget:
- text: Ph
parameters:
negative_prompt: Ph
output:
url: images/Adsฤฑz.png
base_model: stabilityai/stable-diffusion-xl-base-1.0
instance_prompt: Hubble Space Photograph
license: unknown
---
# Hubble Space Photograph
<Gallery />
## Trigger words
You should use `Hubble Space Photograph` to trigger the image generation.
## Download model
Weights for this model are available in Safetensors format.
[Download](/Grekkla/HubbleSpacePhotograph/tree/main) them in the Files & versions tab.
|
LanguageBind/MoE-LLaVA-Phi2-2.7B-4e | LanguageBind | 2024-02-01T07:10:04Z | 389 | 38 | transformers | [
"transformers",
"safetensors",
"moe_llava_phi",
"text-generation",
"custom_code",
"arxiv:2401.15947",
"arxiv:2311.10122",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-23T06:50:28Z | ---
license: apache-2.0
---
<p align="center">
<img src="https://s11.ax1x.com/2023/12/28/piqvDMV.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/abs/2401.15947">MoE-LLaVA: Mixture of Experts for Large Vision-Language Models</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
<h5 align="center">
</h5>
## ๐ฐ News
* **[2024.01.30]** The [paper](https://arxiv.org/abs/2401.15947) is released.
* **[2024.01.27]** ๐ค[Hugging Face demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) and **all codes & datasets** are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
MoE-LLaVA shows excellent performance in multi-modal learning.
### ๐ฅ High performance, but with fewer parameters
- with just **3B sparsely activated parameters**, MoE-LLaVA demonstrates performance comparable to the LLaVA-1.5-7B on various visual understanding datasets and even surpasses the LLaVA-1.5-13B in object hallucination benchmarks.
### ๐ Simple baseline, learning multi-modal interactions with sparse pathways.
- With the addition of **a simple MoE tuning stage**, we can complete the training of MoE-LLaVA on **8 V100 GPUs** within 2 days.
## ๐ค Demo
### Gradio Web UI
Highly recommend trying out our web demo by the following command, which incorporates all features currently supported by MoE-LLaVA. We also provide [online demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) in Huggingface Spaces.
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e"
# use qwen
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e"
# use stablelm
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e"
```
### CLI Inference
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e" --image-file "image.jpg"
# use qwen
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e" --image-file "image.jpg"
# use stablelm
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e" --image-file "image.jpg"
```
## ๐ณ Model Zoo
| Model | LLM | Checkpoint | Avg | VQAv2 | GQA | VizWiz | SQA | T-VQA | POPE | MM-Bench| LLaVA-Bench-Wild | MM-Vet |
|----------|-----------|-----------|---|---|---|---|---|---|---|---|---|---|
| MoE-LLaVA-1.6Bร4-Top2 | 1.6B | [LanguageBind/MoE-LLaVA-StableLM-1.6B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-StableLM-1.6B-4e) | 60.0 | 76.0 | 60.4 | 37.2 | 62.6 | 47.8 | 84.3 | 59.4 | 85.9 | 26.1 |
| MoE-LLaVA-1.8Bร4-Top2 | 1.8B | [LanguageBind/MoE-LLaVA-Qwen-1.8B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Qwen-1.8B-4e) | 60.2 | 76.2 | 61.5 | 32.6 | 63.1 | 48.0 | 87.0 | 59.6 | 88.7 | 25.3 |
| MoE-LLaVA-2.7Bร4-Top2 | 2.7B | [LanguageBind/MoE-LLaVA-Phi2-2.7B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Phi2-2.7B-4e) | 63.9 | 77.1 | 61.1 | 43.4 | 68.7 | 50.2 | 85.0 | 65.5 | 93.2 | 31.1 |
<!--
| LLaVA-1.5 | 7B | [liuhaotian/llava-v1.5-7b](https://huggingface.co/liuhaotian/llava-v1.5-7b) | 62.0 | 78.5 | 62.0 | 50.0 | 66.8 | 58.2 | 85.9 | 64.3 | 31.1 |
| LLaVA-1.5 | 13B | [liuhaotian/llava-v1.5-13b](https://huggingface.co/liuhaotian/llava-v1.5-13b) | 64.9 | 80.0 | 63.3 | 53.6 | 71.6 | 61.3 | 85.9 | 67.7 | 36.1 |
-->
## โ๏ธ Requirements and Installation
* Python >= 3.10
* Pytorch == 2.0.1
* CUDA Version >= 11.7
* **Transformers == 4.36.2**
* **Tokenizers==0.15.1**
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/MoE-LLaVA
cd MoE-LLaVA
conda create -n moellava python=3.10 -y
conda activate moellava
pip install --upgrade pip # enable PEP 660 support
pip install -e .
pip install -e ".[train]"
pip install flash-attn --no-build-isolation
# Below are optional. For Qwen model.
git clone https://github.com/Dao-AILab/flash-attention
cd flash-attention && pip install .
# Below are optional. Installing them might be slow.
# pip install csrc/layer_norm
# If the version of flash-attn is higher than 2.1.1, the following is not needed.
# pip install csrc/rotary
```
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN.md](docs/TRAIN.md) & [EVAL.md](docs/EVAL.md).
## ๐ก Customizing your MoE-LLaVA
The instruction is in [CUSTOM.md](docs/CUSTOM.md).
## ๐ Visualization
The instruction is in [VISUALIZATION.md](docs/VISUALIZATION.md).
## ๐ค API
**We open source all codes.** If you want to load the model (e.g. ```LanguageBind/MoE-LLaVA```) on local, you can use the following code snippets.
**Using the following command to run the code.**
```bash
deepspeed predict.py
```
```python
import torch
from moellava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
from moellava.conversation import conv_templates, SeparatorStyle
from moellava.model.builder import load_pretrained_model
from moellava.utils import disable_torch_init
from moellava.mm_utils import tokenizer_image_token, get_model_name_from_path, KeywordsStoppingCriteria
def main():
disable_torch_init()
image = 'moellava/serve/examples/extreme_ironing.jpg'
inp = 'What is unusual about this image?'
model_path = 'LanguageBind/MoE-LLaVA-Phi2-2.7B-4e' # LanguageBind/MoE-LLaVA-Qwen-1.8B-4e or LanguageBind/MoE-LLaVA-StableLM-1.6B-4e
device = 'cuda'
load_4bit, load_8bit = False, False # FIXME: Deepspeed support 4bit or 8bit?
model_name = get_model_name_from_path(model_path)
tokenizer, model, processor, context_len = load_pretrained_model(model_path, None, model_name, load_8bit, load_4bit, device=device)
image_processor = processor['image']
conv_mode = "phi" # qwen or stablelm
conv = conv_templates[conv_mode].copy()
roles = conv.roles
image_tensor = image_processor.preprocess(image, return_tensors='pt')['pixel_values'].to(model.device, dtype=torch.float16)
print(f"{roles[1]}: {inp}")
inp = DEFAULT_IMAGE_TOKEN + '\n' + inp
conv.append_message(conv.roles[0], inp)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).cuda()
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=image_tensor,
do_sample=True,
temperature=0.2,
max_new_tokens=1024,
use_cache=True,
stopping_criteria=[stopping_criteria])
outputs = tokenizer.decode(output_ids[0, input_ids.shape[1]:], skip_special_tokens=True).strip()
print(outputs)
if __name__ == '__main__':
main()
```
## ๐ Related Projects
* [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) This framework empowers the model to efficiently utilize the united visual tokens.
* [LanguageBind](https://github.com/PKU-YuanGroup/LanguageBind) An open source five modalities language-based retrieval framework.
## ๐ Acknowledgement
* [LLaVA](https://github.com/haotian-liu/LLaVA) The codebase we built upon and it is an efficient large language and vision assistant.
## ๐ License
* The majority of this project is released under the Apache 2.0 license as found in the [LICENSE](https://github.com/PKU-YuanGroup/MoE-LLaVA/blob/main/LICENSE) file.
* The service is a research preview intended for non-commercial use only, subject to the model [License](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md) of LLaMA, [Terms of Use](https://openai.com/policies/terms-of-use) of the data generated by OpenAI, and [Privacy Practices](https://chrome.google.com/webstore/detail/sharegpt-share-your-chatg/daiacboceoaocpibfodeljbdfacokfjb) of ShareGPT. Please contact us if you find any potential violation.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{lin2024moellava,
title={MoE-LLaVA: Mixture of Experts for Large Vision-Language Models},
author={Bin Lin and Zhenyu Tang and Yang Ye and Jiaxi Cui and Bin Zhu and Peng Jin and Junwu Zhang and Munan Ning and Li Yuan},
year={2024},
eprint={2401.15947},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
```BibTeX
@article{lin2023video,
title={Video-LLaVA: Learning United Visual Representation by Alignment Before Projection},
author={Lin, Bin and Zhu, Bin and Ye, Yang and Ning, Munan and Jin, Peng and Yuan, Li},
journal={arXiv preprint arXiv:2311.10122},
year={2023}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/MoE-LLaVA&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/MoE-LLaVA/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/MoE-LLaVA" />
</a> |
LanguageBind/LanguageBind_Video_V1.5_FT | LanguageBind | 2024-02-01T06:58:17Z | 1,905 | 4 | transformers | [
"transformers",
"pytorch",
"LanguageBindVideo",
"zero-shot-image-classification",
"arxiv:2310.01852",
"license:mit",
"endpoints_compatible",
"region:us"
]
| zero-shot-image-classification | 2023-11-26T13:35:39Z | ---
license: mit
---
<p align="center">
<img src="https://s11.ax1x.com/2024/02/01/pFMDAm9.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/pdf/2310.01852.pdf">ใICLR 2024 ๐ฅใLanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
## ๐ฐ News
* **[2024.01.27]** ๐๐๐ Our [MoE-LLaVA](https://github.com/PKU-YuanGroup/MoE-LLaVA) is released! A sparse model with 3B parameters outperformed the dense model with 7B parameters.
* **[2024.01.16]** ๐ฅ๐ฅ๐ฅ Our LanguageBind has been accepted at ICLR 2024! We earn the score of 6(3)8(6)6(6)6(6) [here](https://openreview.net/forum?id=QmZKc7UZCy¬eId=OgsxQxAleA).
* **[2023.12.15]** ๐ช๐ช๐ช We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M video-text data**. We launch **LanguageBind_Video 1.5**, checking our [model zoo](#-model-zoo).
* **[2023.12.10]** We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M depth and 10M thermal data**. We are in the process of uploading thermal and depth data on [Hugging Face](https://huggingface.co/datasets/LanguageBind/VIDAL-Depth-Thermal) and expect the whole process to last 1-2 months.
* **[2023.11.27]** ๐ฅ๐ฅ๐ฅ We have updated our [paper](https://arxiv.org/abs/2310.01852) with emergency zero-shot results., checking our โจ [results](#emergency-results).
* **[2023.11.26]** ๐ฅ๐ฅ๐ฅ We have open-sourced all textual sources and corresponding YouTube IDs [here](DATASETS.md).
* **[2023.11.26]** ๐ฃ๐ฃ๐ฃ We have open-sourced fully fine-tuned **Video & Audio**, achieving improved performance once again, checking our [model zoo](#-model-zoo).
* **[2023.11.22]** We are about to release a fully fine-tuned version, and the **HUGE** version is currently undergoing training.
* **[2023.11.21]** ๐ฅ We are releasing sample data in [DATASETS.md](DATASETS.md) so that individuals who are interested can further modify the code to train it on their own data.
* **[2023.11.20]** ๐๐๐ [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) builds a large visual-language model to achieve ๐SOTA performances based on LanguageBind encoders.
* **[2023.10.23]** ๐ถ LanguageBind-Audio achieves ๐๐๐**state-of-the-art (SOTA) performance on 5 datasets**, checking our โจ [results](#multiple-modalities)!
* **[2023.10.14]** ๐ฑ Released a stronger LanguageBind-Video, checking our โจ [results](#video-language)! The video checkpoint **have updated** on Huggingface Model Hub!
* **[2023.10.10]** We provide sample data, which can be found in [assets](assets), and [emergency zero-shot usage](#emergency-zero-shot) is described.
* **[2023.10.07]** The checkpoints are available on ๐ค [Huggingface Model](https://huggingface.co/LanguageBind).
* **[2023.10.04]** Code and [demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
### ๐ก High performance, but NO intermediate modality required
LanguageBind is a **language-centric** multimodal pretraining approach, **taking the language as the bind across different modalities** because the language modality is well-explored and contains rich semantics.
* The following first figure shows the architecture of LanguageBind. LanguageBind can be easily extended to segmentation, detection tasks, and potentially to unlimited modalities.
### โก๏ธ A multimodal, fully aligned and voluminous dataset
We propose **VIDAL-10M**, **10 Million data** with **V**ideo, **I**nfrared, **D**epth, **A**udio and their corresponding **L**anguage, which greatly expands the data beyond visual modalities.
* The second figure shows our proposed VIDAL-10M dataset, which includes five modalities: video, infrared, depth, audio, and language.
### ๐ฅ Multi-view enhanced description for training
We make multi-view enhancements to language. We produce multi-view description that combines **meta-data**, **spatial**, and **temporal** to greatly enhance the semantic information of the language. In addition we further **enhance the language with ChatGPT** to create a good semantic space for each modality aligned language.
## ๐ค Demo
* **Local demo.** Highly recommend trying out our web demo, which incorporates all features currently supported by LanguageBind.
```bash
python gradio_app.py
```
* **Online demo.** We provide the [online demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) in Huggingface Spaces. In this demo, you can calculate the similarity of modalities to language, such as audio-to-language, video-to-language, and depth-to-image.
## ๐ ๏ธ Requirements and Installation
* Python >= 3.8
* Pytorch >= 1.13.1
* CUDA Version >= 11.6
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/LanguageBind
cd LanguageBind
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116
pip install -r requirements.txt
```
## ๐ณ Model Zoo
The names in the table represent different encoder models. For example, `LanguageBind/LanguageBind_Video_FT` represents the fully fine-tuned version, while `LanguageBind/LanguageBind_Video` represents the LoRA-tuned version.
You can freely replace them in the recommended [API usage](#-api). We recommend using the fully fine-tuned version, as it offers stronger performance.
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Modality</th><th>LoRA tuning</th><th>Fine-tuning</th>
</tr>
<tr align="center">
<td>Video</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">LanguageBind_Video</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">LanguageBind_Video_FT</a></td>
</tr>
<tr align="center">
<td>Audio</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio">LanguageBind_Audio</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio_FT">LanguageBind_Audio_FT</a></td>
</tr>
<tr align="center">
<td>Depth</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Depth">LanguageBind_Depth</a></td><td>-</td>
</tr>
<tr align="center">
<td>Thermal</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Thermal">LanguageBind_Thermal</a></td><td>-</td>
</tr>
</table>
</div>
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Version</th><th>Tuning</th><th>Model size</th><th>Num_frames</th><th>HF Link</th><th>MSR-VTT</th><th>DiDeMo</th><th>ActivityNet</th><th>MSVD</th>
</tr>
<tr align="center">
<td>LanguageBind_Video</td><td>LoRA</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">Link</a></td><td>42.6</td><td>37.8</td><td>35.1</td><td>52.2</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">Link</a></td><td>42.7</td><td>38.1</td><td>36.9</td><td>53.5</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_V1.5_FT">Link</a></td><td>42.8</td><td>39.7</td><td>38.4</td><td>54.1</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>12</td><td>Coming soon</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_Huge_V1.5_FT">Link</a></td><td>44.8</td><td>39.9</td><td>41.0</td><td>53.7</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>12</td><td>Coming soon</td>
</tr>
</table>
</div>
## ๐ค API
**We open source all modalities preprocessing code.** If you want to load the model (e.g. ```LanguageBind/LanguageBind_Thermal```) from the model hub on Huggingface or on local, you can use the following code snippets!
### Inference for Multi-modal Binding
We have provided some sample datasets in [assets](assets) to quickly see how languagebind works.
```python
import torch
from languagebind import LanguageBind, to_device, transform_dict, LanguageBindImageTokenizer
if __name__ == '__main__':
device = 'cuda:0'
device = torch.device(device)
clip_type = {
'video': 'LanguageBind_Video_FT', # also LanguageBind_Video
'audio': 'LanguageBind_Audio_FT', # also LanguageBind_Audio
'thermal': 'LanguageBind_Thermal',
'image': 'LanguageBind_Image',
'depth': 'LanguageBind_Depth',
}
model = LanguageBind(clip_type=clip_type, cache_dir='./cache_dir')
model = model.to(device)
model.eval()
pretrained_ckpt = f'lb203/LanguageBind_Image'
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir/tokenizer_cache_dir')
modality_transform = {c: transform_dict[c](model.modality_config[c]) for c in clip_type.keys()}
image = ['assets/image/0.jpg', 'assets/image/1.jpg']
audio = ['assets/audio/0.wav', 'assets/audio/1.wav']
video = ['assets/video/0.mp4', 'assets/video/1.mp4']
depth = ['assets/depth/0.png', 'assets/depth/1.png']
thermal = ['assets/thermal/0.jpg', 'assets/thermal/1.jpg']
language = ["Training a parakeet to climb up a ladder.", 'A lion climbing a tree to catch a monkey.']
inputs = {
'image': to_device(modality_transform['image'](image), device),
'video': to_device(modality_transform['video'](video), device),
'audio': to_device(modality_transform['audio'](audio), device),
'depth': to_device(modality_transform['depth'](depth), device),
'thermal': to_device(modality_transform['thermal'](thermal), device),
}
inputs['language'] = to_device(tokenizer(language, max_length=77, padding='max_length',
truncation=True, return_tensors='pt'), device)
with torch.no_grad():
embeddings = model(inputs)
print("Video x Text: \n",
torch.softmax(embeddings['video'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Image x Text: \n",
torch.softmax(embeddings['image'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Depth x Text: \n",
torch.softmax(embeddings['depth'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Audio x Text: \n",
torch.softmax(embeddings['audio'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Thermal x Text: \n",
torch.softmax(embeddings['thermal'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
```
Then returns the following result.
```bash
Video x Text:
[[9.9989331e-01 1.0667283e-04]
[1.3255903e-03 9.9867439e-01]]
Image x Text:
[[9.9990666e-01 9.3292067e-05]
[4.6132666e-08 1.0000000e+00]]
Depth x Text:
[[0.9954276 0.00457235]
[0.12042473 0.8795753 ]]
Audio x Text:
[[0.97634876 0.02365119]
[0.02917843 0.97082156]]
Thermal x Text:
[[0.9482511 0.0517489 ]
[0.48746133 0.5125386 ]]
```
### Emergency zero-shot
Since languagebind binds each modality together, we also found the **emergency zero-shot**. It's very simple to use.
```python
print("Video x Audio: \n", torch.softmax(embeddings['video'] @ embeddings['audio'].T, dim=-1).detach().cpu().numpy())
print("Image x Depth: \n", torch.softmax(embeddings['image'] @ embeddings['depth'].T, dim=-1).detach().cpu().numpy())
print("Image x Thermal: \n", torch.softmax(embeddings['image'] @ embeddings['thermal'].T, dim=-1).detach().cpu().numpy())
```
Then, you will get:
```
Video x Audio:
[[1.0000000e+00 0.0000000e+00]
[3.1150486e-32 1.0000000e+00]]
Image x Depth:
[[1. 0.]
[0. 1.]]
Image x Thermal:
[[1. 0.]
[0. 1.]]
```
### Different branches for X-Language task
Additionally, LanguageBind can be **disassembled into different branches** to handle different tasks. Note that we do not train Image, which just initialize from OpenCLIP.
#### Thermal
```python
import torch
from languagebind import LanguageBindThermal, LanguageBindThermalTokenizer, LanguageBindThermalProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Thermal'
model = LanguageBindThermal.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindThermalTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
thermal_process = LanguageBindThermalProcessor(model.config, tokenizer)
model.eval()
data = thermal_process([r"your/thermal.jpg"], ['your text'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Depth
```python
import torch
from languagebind import LanguageBindDepth, LanguageBindDepthTokenizer, LanguageBindDepthProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Depth'
model = LanguageBindDepth.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindDepthTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
depth_process = LanguageBindDepthProcessor(model.config, tokenizer)
model.eval()
data = depth_process([r"your/depth.png"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Video
```python
import torch
from languagebind import LanguageBindVideo, LanguageBindVideoTokenizer, LanguageBindVideoProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Video_FT' # also 'LanguageBind/LanguageBind_Video'
model = LanguageBindVideo.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindVideoTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
video_process = LanguageBindVideoProcessor(model.config, tokenizer)
model.eval()
data = video_process(["your/video.mp4"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Audio
```python
import torch
from languagebind import LanguageBindAudio, LanguageBindAudioTokenizer, LanguageBindAudioProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Audio_FT' # also 'LanguageBind/LanguageBind_Audio'
model = LanguageBindAudio.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindAudioTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
audio_process = LanguageBindAudioProcessor(model.config, tokenizer)
model.eval()
data = audio_process([r"your/audio.wav"], ['your audio.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Image
Note that our image encoder is the same as OpenCLIP. **Not** as fine-tuned as other modalities.
```python
import torch
from languagebind import LanguageBindImage, LanguageBindImageTokenizer, LanguageBindImageProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Image'
model = LanguageBindImage.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
image_process = LanguageBindImageProcessor(model.config, tokenizer)
model.eval()
data = image_process([r"your/image.jpg"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
## ๐ฅ VIDAL-10M
The datasets is in [DATASETS.md](DATASETS.md).
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN_AND_VALIDATE.md](TRAIN_AND_VALIDATE.md).
## ๐ Acknowledgement
* [OpenCLIP](https://github.com/mlfoundations/open_clip) An open source pretraining framework.
* [CLIP4Clip](https://github.com/ArrowLuo/CLIP4Clip) An open source Video-Text retrieval framework.
* [sRGB-TIR](https://github.com/rpmsnu/sRGB-TIR) An open source framework to generate infrared (thermal) images.
* [GLPN](https://github.com/vinvino02/GLPDepth) An open source framework to generate depth images.
## ๐ License
* The majority of this project is released under the MIT license as found in the [LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/LICENSE) file.
* The dataset of this project is released under the CC-BY-NC 4.0 license as found in the [DATASET_LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/DATASET_LICENSE) file.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{zhu2023languagebind,
title={LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment},
author={Bin Zhu and Bin Lin and Munan Ning and Yang Yan and Jiaxi Cui and Wang HongFa and Yatian Pang and Wenhao Jiang and Junwu Zhang and Zongwei Li and Cai Wan Zhang and Zhifeng Li and Wei Liu and Li Yuan},
year={2023},
eprint={2310.01852},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/LanguageBind&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/LanguageBind/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/LanguageBind" />
</a>
|
LanguageBind/LanguageBind_Audio_V1.5 | LanguageBind | 2024-02-01T06:58:03Z | 0 | 1 | null | [
"arxiv:2310.01852",
"license:mit",
"region:us"
]
| null | 2023-11-26T13:36:02Z | ---
license: mit
---
<p align="center">
<img src="https://s11.ax1x.com/2024/02/01/pFMDAm9.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/pdf/2310.01852.pdf">ใICLR 2024 ๐ฅใLanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
## ๐ฐ News
* **[2024.01.27]** ๐๐๐ Our [MoE-LLaVA](https://github.com/PKU-YuanGroup/MoE-LLaVA) is released! A sparse model with 3B parameters outperformed the dense model with 7B parameters.
* **[2024.01.16]** ๐ฅ๐ฅ๐ฅ Our LanguageBind has been accepted at ICLR 2024! We earn the score of 6(3)8(6)6(6)6(6) [here](https://openreview.net/forum?id=QmZKc7UZCy¬eId=OgsxQxAleA).
* **[2023.12.15]** ๐ช๐ช๐ช We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M video-text data**. We launch **LanguageBind_Video 1.5**, checking our [model zoo](#-model-zoo).
* **[2023.12.10]** We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M depth and 10M thermal data**. We are in the process of uploading thermal and depth data on [Hugging Face](https://huggingface.co/datasets/LanguageBind/VIDAL-Depth-Thermal) and expect the whole process to last 1-2 months.
* **[2023.11.27]** ๐ฅ๐ฅ๐ฅ We have updated our [paper](https://arxiv.org/abs/2310.01852) with emergency zero-shot results., checking our โจ [results](#emergency-results).
* **[2023.11.26]** ๐ฅ๐ฅ๐ฅ We have open-sourced all textual sources and corresponding YouTube IDs [here](DATASETS.md).
* **[2023.11.26]** ๐ฃ๐ฃ๐ฃ We have open-sourced fully fine-tuned **Video & Audio**, achieving improved performance once again, checking our [model zoo](#-model-zoo).
* **[2023.11.22]** We are about to release a fully fine-tuned version, and the **HUGE** version is currently undergoing training.
* **[2023.11.21]** ๐ฅ We are releasing sample data in [DATASETS.md](DATASETS.md) so that individuals who are interested can further modify the code to train it on their own data.
* **[2023.11.20]** ๐๐๐ [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) builds a large visual-language model to achieve ๐SOTA performances based on LanguageBind encoders.
* **[2023.10.23]** ๐ถ LanguageBind-Audio achieves ๐๐๐**state-of-the-art (SOTA) performance on 5 datasets**, checking our โจ [results](#multiple-modalities)!
* **[2023.10.14]** ๐ฑ Released a stronger LanguageBind-Video, checking our โจ [results](#video-language)! The video checkpoint **have updated** on Huggingface Model Hub!
* **[2023.10.10]** We provide sample data, which can be found in [assets](assets), and [emergency zero-shot usage](#emergency-zero-shot) is described.
* **[2023.10.07]** The checkpoints are available on ๐ค [Huggingface Model](https://huggingface.co/LanguageBind).
* **[2023.10.04]** Code and [demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
### ๐ก High performance, but NO intermediate modality required
LanguageBind is a **language-centric** multimodal pretraining approach, **taking the language as the bind across different modalities** because the language modality is well-explored and contains rich semantics.
* The following first figure shows the architecture of LanguageBind. LanguageBind can be easily extended to segmentation, detection tasks, and potentially to unlimited modalities.
### โก๏ธ A multimodal, fully aligned and voluminous dataset
We propose **VIDAL-10M**, **10 Million data** with **V**ideo, **I**nfrared, **D**epth, **A**udio and their corresponding **L**anguage, which greatly expands the data beyond visual modalities.
* The second figure shows our proposed VIDAL-10M dataset, which includes five modalities: video, infrared, depth, audio, and language.
### ๐ฅ Multi-view enhanced description for training
We make multi-view enhancements to language. We produce multi-view description that combines **meta-data**, **spatial**, and **temporal** to greatly enhance the semantic information of the language. In addition we further **enhance the language with ChatGPT** to create a good semantic space for each modality aligned language.
## ๐ค Demo
* **Local demo.** Highly recommend trying out our web demo, which incorporates all features currently supported by LanguageBind.
```bash
python gradio_app.py
```
* **Online demo.** We provide the [online demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) in Huggingface Spaces. In this demo, you can calculate the similarity of modalities to language, such as audio-to-language, video-to-language, and depth-to-image.
## ๐ ๏ธ Requirements and Installation
* Python >= 3.8
* Pytorch >= 1.13.1
* CUDA Version >= 11.6
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/LanguageBind
cd LanguageBind
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116
pip install -r requirements.txt
```
## ๐ณ Model Zoo
The names in the table represent different encoder models. For example, `LanguageBind/LanguageBind_Video_FT` represents the fully fine-tuned version, while `LanguageBind/LanguageBind_Video` represents the LoRA-tuned version.
You can freely replace them in the recommended [API usage](#-api). We recommend using the fully fine-tuned version, as it offers stronger performance.
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Modality</th><th>LoRA tuning</th><th>Fine-tuning</th>
</tr>
<tr align="center">
<td>Video</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">LanguageBind_Video</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">LanguageBind_Video_FT</a></td>
</tr>
<tr align="center">
<td>Audio</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio">LanguageBind_Audio</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio_FT">LanguageBind_Audio_FT</a></td>
</tr>
<tr align="center">
<td>Depth</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Depth">LanguageBind_Depth</a></td><td>-</td>
</tr>
<tr align="center">
<td>Thermal</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Thermal">LanguageBind_Thermal</a></td><td>-</td>
</tr>
</table>
</div>
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Version</th><th>Tuning</th><th>Model size</th><th>Num_frames</th><th>HF Link</th><th>MSR-VTT</th><th>DiDeMo</th><th>ActivityNet</th><th>MSVD</th>
</tr>
<tr align="center">
<td>LanguageBind_Video</td><td>LoRA</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">Link</a></td><td>42.6</td><td>37.8</td><td>35.1</td><td>52.2</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">Link</a></td><td>42.7</td><td>38.1</td><td>36.9</td><td>53.5</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_V1.5_FT">Link</a></td><td>42.8</td><td>39.7</td><td>38.4</td><td>54.1</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>12</td><td>Coming soon</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_Huge_V1.5_FT">Link</a></td><td>44.8</td><td>39.9</td><td>41.0</td><td>53.7</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>12</td><td>Coming soon</td>
</tr>
</table>
</div>
## ๐ค API
**We open source all modalities preprocessing code.** If you want to load the model (e.g. ```LanguageBind/LanguageBind_Thermal```) from the model hub on Huggingface or on local, you can use the following code snippets!
### Inference for Multi-modal Binding
We have provided some sample datasets in [assets](assets) to quickly see how languagebind works.
```python
import torch
from languagebind import LanguageBind, to_device, transform_dict, LanguageBindImageTokenizer
if __name__ == '__main__':
device = 'cuda:0'
device = torch.device(device)
clip_type = {
'video': 'LanguageBind_Video_FT', # also LanguageBind_Video
'audio': 'LanguageBind_Audio_FT', # also LanguageBind_Audio
'thermal': 'LanguageBind_Thermal',
'image': 'LanguageBind_Image',
'depth': 'LanguageBind_Depth',
}
model = LanguageBind(clip_type=clip_type, cache_dir='./cache_dir')
model = model.to(device)
model.eval()
pretrained_ckpt = f'lb203/LanguageBind_Image'
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir/tokenizer_cache_dir')
modality_transform = {c: transform_dict[c](model.modality_config[c]) for c in clip_type.keys()}
image = ['assets/image/0.jpg', 'assets/image/1.jpg']
audio = ['assets/audio/0.wav', 'assets/audio/1.wav']
video = ['assets/video/0.mp4', 'assets/video/1.mp4']
depth = ['assets/depth/0.png', 'assets/depth/1.png']
thermal = ['assets/thermal/0.jpg', 'assets/thermal/1.jpg']
language = ["Training a parakeet to climb up a ladder.", 'A lion climbing a tree to catch a monkey.']
inputs = {
'image': to_device(modality_transform['image'](image), device),
'video': to_device(modality_transform['video'](video), device),
'audio': to_device(modality_transform['audio'](audio), device),
'depth': to_device(modality_transform['depth'](depth), device),
'thermal': to_device(modality_transform['thermal'](thermal), device),
}
inputs['language'] = to_device(tokenizer(language, max_length=77, padding='max_length',
truncation=True, return_tensors='pt'), device)
with torch.no_grad():
embeddings = model(inputs)
print("Video x Text: \n",
torch.softmax(embeddings['video'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Image x Text: \n",
torch.softmax(embeddings['image'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Depth x Text: \n",
torch.softmax(embeddings['depth'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Audio x Text: \n",
torch.softmax(embeddings['audio'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Thermal x Text: \n",
torch.softmax(embeddings['thermal'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
```
Then returns the following result.
```bash
Video x Text:
[[9.9989331e-01 1.0667283e-04]
[1.3255903e-03 9.9867439e-01]]
Image x Text:
[[9.9990666e-01 9.3292067e-05]
[4.6132666e-08 1.0000000e+00]]
Depth x Text:
[[0.9954276 0.00457235]
[0.12042473 0.8795753 ]]
Audio x Text:
[[0.97634876 0.02365119]
[0.02917843 0.97082156]]
Thermal x Text:
[[0.9482511 0.0517489 ]
[0.48746133 0.5125386 ]]
```
### Emergency zero-shot
Since languagebind binds each modality together, we also found the **emergency zero-shot**. It's very simple to use.
```python
print("Video x Audio: \n", torch.softmax(embeddings['video'] @ embeddings['audio'].T, dim=-1).detach().cpu().numpy())
print("Image x Depth: \n", torch.softmax(embeddings['image'] @ embeddings['depth'].T, dim=-1).detach().cpu().numpy())
print("Image x Thermal: \n", torch.softmax(embeddings['image'] @ embeddings['thermal'].T, dim=-1).detach().cpu().numpy())
```
Then, you will get:
```
Video x Audio:
[[1.0000000e+00 0.0000000e+00]
[3.1150486e-32 1.0000000e+00]]
Image x Depth:
[[1. 0.]
[0. 1.]]
Image x Thermal:
[[1. 0.]
[0. 1.]]
```
### Different branches for X-Language task
Additionally, LanguageBind can be **disassembled into different branches** to handle different tasks. Note that we do not train Image, which just initialize from OpenCLIP.
#### Thermal
```python
import torch
from languagebind import LanguageBindThermal, LanguageBindThermalTokenizer, LanguageBindThermalProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Thermal'
model = LanguageBindThermal.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindThermalTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
thermal_process = LanguageBindThermalProcessor(model.config, tokenizer)
model.eval()
data = thermal_process([r"your/thermal.jpg"], ['your text'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Depth
```python
import torch
from languagebind import LanguageBindDepth, LanguageBindDepthTokenizer, LanguageBindDepthProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Depth'
model = LanguageBindDepth.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindDepthTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
depth_process = LanguageBindDepthProcessor(model.config, tokenizer)
model.eval()
data = depth_process([r"your/depth.png"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Video
```python
import torch
from languagebind import LanguageBindVideo, LanguageBindVideoTokenizer, LanguageBindVideoProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Video_FT' # also 'LanguageBind/LanguageBind_Video'
model = LanguageBindVideo.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindVideoTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
video_process = LanguageBindVideoProcessor(model.config, tokenizer)
model.eval()
data = video_process(["your/video.mp4"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Audio
```python
import torch
from languagebind import LanguageBindAudio, LanguageBindAudioTokenizer, LanguageBindAudioProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Audio_FT' # also 'LanguageBind/LanguageBind_Audio'
model = LanguageBindAudio.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindAudioTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
audio_process = LanguageBindAudioProcessor(model.config, tokenizer)
model.eval()
data = audio_process([r"your/audio.wav"], ['your audio.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Image
Note that our image encoder is the same as OpenCLIP. **Not** as fine-tuned as other modalities.
```python
import torch
from languagebind import LanguageBindImage, LanguageBindImageTokenizer, LanguageBindImageProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Image'
model = LanguageBindImage.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
image_process = LanguageBindImageProcessor(model.config, tokenizer)
model.eval()
data = image_process([r"your/image.jpg"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
## ๐ฅ VIDAL-10M
The datasets is in [DATASETS.md](DATASETS.md).
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN_AND_VALIDATE.md](TRAIN_AND_VALIDATE.md).
## ๐ Acknowledgement
* [OpenCLIP](https://github.com/mlfoundations/open_clip) An open source pretraining framework.
* [CLIP4Clip](https://github.com/ArrowLuo/CLIP4Clip) An open source Video-Text retrieval framework.
* [sRGB-TIR](https://github.com/rpmsnu/sRGB-TIR) An open source framework to generate infrared (thermal) images.
* [GLPN](https://github.com/vinvino02/GLPDepth) An open source framework to generate depth images.
## ๐ License
* The majority of this project is released under the MIT license as found in the [LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/LICENSE) file.
* The dataset of this project is released under the CC-BY-NC 4.0 license as found in the [DATASET_LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/DATASET_LICENSE) file.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{zhu2023languagebind,
title={LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment},
author={Bin Zhu and Bin Lin and Munan Ning and Yang Yan and Jiaxi Cui and Wang HongFa and Yatian Pang and Wenhao Jiang and Junwu Zhang and Zongwei Li and Cai Wan Zhang and Zhifeng Li and Wei Liu and Li Yuan},
year={2023},
eprint={2310.01852},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/LanguageBind&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/LanguageBind/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/LanguageBind" />
</a>
|
LanguageBind/LanguageBind_Video | LanguageBind | 2024-02-01T06:57:36Z | 364 | 2 | transformers | [
"transformers",
"pytorch",
"LanguageBindVideo",
"zero-shot-image-classification",
"arxiv:2310.01852",
"license:mit",
"endpoints_compatible",
"region:us"
]
| zero-shot-image-classification | 2023-10-06T09:07:15Z | ---
license: mit
---
<p align="center">
<img src="https://s11.ax1x.com/2024/02/01/pFMDAm9.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/pdf/2310.01852.pdf">ใICLR 2024 ๐ฅใLanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
## ๐ฐ News
* **[2024.01.27]** ๐๐๐ Our [MoE-LLaVA](https://github.com/PKU-YuanGroup/MoE-LLaVA) is released! A sparse model with 3B parameters outperformed the dense model with 7B parameters.
* **[2024.01.16]** ๐ฅ๐ฅ๐ฅ Our LanguageBind has been accepted at ICLR 2024! We earn the score of 6(3)8(6)6(6)6(6) [here](https://openreview.net/forum?id=QmZKc7UZCy¬eId=OgsxQxAleA).
* **[2023.12.15]** ๐ช๐ช๐ช We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M video-text data**. We launch **LanguageBind_Video 1.5**, checking our [model zoo](#-model-zoo).
* **[2023.12.10]** We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M depth and 10M thermal data**. We are in the process of uploading thermal and depth data on [Hugging Face](https://huggingface.co/datasets/LanguageBind/VIDAL-Depth-Thermal) and expect the whole process to last 1-2 months.
* **[2023.11.27]** ๐ฅ๐ฅ๐ฅ We have updated our [paper](https://arxiv.org/abs/2310.01852) with emergency zero-shot results., checking our โจ [results](#emergency-results).
* **[2023.11.26]** ๐ฅ๐ฅ๐ฅ We have open-sourced all textual sources and corresponding YouTube IDs [here](DATASETS.md).
* **[2023.11.26]** ๐ฃ๐ฃ๐ฃ We have open-sourced fully fine-tuned **Video & Audio**, achieving improved performance once again, checking our [model zoo](#-model-zoo).
* **[2023.11.22]** We are about to release a fully fine-tuned version, and the **HUGE** version is currently undergoing training.
* **[2023.11.21]** ๐ฅ We are releasing sample data in [DATASETS.md](DATASETS.md) so that individuals who are interested can further modify the code to train it on their own data.
* **[2023.11.20]** ๐๐๐ [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) builds a large visual-language model to achieve ๐SOTA performances based on LanguageBind encoders.
* **[2023.10.23]** ๐ถ LanguageBind-Audio achieves ๐๐๐**state-of-the-art (SOTA) performance on 5 datasets**, checking our โจ [results](#multiple-modalities)!
* **[2023.10.14]** ๐ฑ Released a stronger LanguageBind-Video, checking our โจ [results](#video-language)! The video checkpoint **have updated** on Huggingface Model Hub!
* **[2023.10.10]** We provide sample data, which can be found in [assets](assets), and [emergency zero-shot usage](#emergency-zero-shot) is described.
* **[2023.10.07]** The checkpoints are available on ๐ค [Huggingface Model](https://huggingface.co/LanguageBind).
* **[2023.10.04]** Code and [demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
### ๐ก High performance, but NO intermediate modality required
LanguageBind is a **language-centric** multimodal pretraining approach, **taking the language as the bind across different modalities** because the language modality is well-explored and contains rich semantics.
* The following first figure shows the architecture of LanguageBind. LanguageBind can be easily extended to segmentation, detection tasks, and potentially to unlimited modalities.
### โก๏ธ A multimodal, fully aligned and voluminous dataset
We propose **VIDAL-10M**, **10 Million data** with **V**ideo, **I**nfrared, **D**epth, **A**udio and their corresponding **L**anguage, which greatly expands the data beyond visual modalities.
* The second figure shows our proposed VIDAL-10M dataset, which includes five modalities: video, infrared, depth, audio, and language.
### ๐ฅ Multi-view enhanced description for training
We make multi-view enhancements to language. We produce multi-view description that combines **meta-data**, **spatial**, and **temporal** to greatly enhance the semantic information of the language. In addition we further **enhance the language with ChatGPT** to create a good semantic space for each modality aligned language.
## ๐ค Demo
* **Local demo.** Highly recommend trying out our web demo, which incorporates all features currently supported by LanguageBind.
```bash
python gradio_app.py
```
* **Online demo.** We provide the [online demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) in Huggingface Spaces. In this demo, you can calculate the similarity of modalities to language, such as audio-to-language, video-to-language, and depth-to-image.
## ๐ ๏ธ Requirements and Installation
* Python >= 3.8
* Pytorch >= 1.13.1
* CUDA Version >= 11.6
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/LanguageBind
cd LanguageBind
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116
pip install -r requirements.txt
```
## ๐ณ Model Zoo
The names in the table represent different encoder models. For example, `LanguageBind/LanguageBind_Video_FT` represents the fully fine-tuned version, while `LanguageBind/LanguageBind_Video` represents the LoRA-tuned version.
You can freely replace them in the recommended [API usage](#-api). We recommend using the fully fine-tuned version, as it offers stronger performance.
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Modality</th><th>LoRA tuning</th><th>Fine-tuning</th>
</tr>
<tr align="center">
<td>Video</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">LanguageBind_Video</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">LanguageBind_Video_FT</a></td>
</tr>
<tr align="center">
<td>Audio</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio">LanguageBind_Audio</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio_FT">LanguageBind_Audio_FT</a></td>
</tr>
<tr align="center">
<td>Depth</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Depth">LanguageBind_Depth</a></td><td>-</td>
</tr>
<tr align="center">
<td>Thermal</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Thermal">LanguageBind_Thermal</a></td><td>-</td>
</tr>
</table>
</div>
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Version</th><th>Tuning</th><th>Model size</th><th>Num_frames</th><th>HF Link</th><th>MSR-VTT</th><th>DiDeMo</th><th>ActivityNet</th><th>MSVD</th>
</tr>
<tr align="center">
<td>LanguageBind_Video</td><td>LoRA</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">Link</a></td><td>42.6</td><td>37.8</td><td>35.1</td><td>52.2</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">Link</a></td><td>42.7</td><td>38.1</td><td>36.9</td><td>53.5</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_V1.5_FT">Link</a></td><td>42.8</td><td>39.7</td><td>38.4</td><td>54.1</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>12</td><td>Coming soon</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_Huge_V1.5_FT">Link</a></td><td>44.8</td><td>39.9</td><td>41.0</td><td>53.7</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>12</td><td>Coming soon</td>
</tr>
</table>
</div>
## ๐ค API
**We open source all modalities preprocessing code.** If you want to load the model (e.g. ```LanguageBind/LanguageBind_Thermal```) from the model hub on Huggingface or on local, you can use the following code snippets!
### Inference for Multi-modal Binding
We have provided some sample datasets in [assets](assets) to quickly see how languagebind works.
```python
import torch
from languagebind import LanguageBind, to_device, transform_dict, LanguageBindImageTokenizer
if __name__ == '__main__':
device = 'cuda:0'
device = torch.device(device)
clip_type = {
'video': 'LanguageBind_Video_FT', # also LanguageBind_Video
'audio': 'LanguageBind_Audio_FT', # also LanguageBind_Audio
'thermal': 'LanguageBind_Thermal',
'image': 'LanguageBind_Image',
'depth': 'LanguageBind_Depth',
}
model = LanguageBind(clip_type=clip_type, cache_dir='./cache_dir')
model = model.to(device)
model.eval()
pretrained_ckpt = f'lb203/LanguageBind_Image'
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir/tokenizer_cache_dir')
modality_transform = {c: transform_dict[c](model.modality_config[c]) for c in clip_type.keys()}
image = ['assets/image/0.jpg', 'assets/image/1.jpg']
audio = ['assets/audio/0.wav', 'assets/audio/1.wav']
video = ['assets/video/0.mp4', 'assets/video/1.mp4']
depth = ['assets/depth/0.png', 'assets/depth/1.png']
thermal = ['assets/thermal/0.jpg', 'assets/thermal/1.jpg']
language = ["Training a parakeet to climb up a ladder.", 'A lion climbing a tree to catch a monkey.']
inputs = {
'image': to_device(modality_transform['image'](image), device),
'video': to_device(modality_transform['video'](video), device),
'audio': to_device(modality_transform['audio'](audio), device),
'depth': to_device(modality_transform['depth'](depth), device),
'thermal': to_device(modality_transform['thermal'](thermal), device),
}
inputs['language'] = to_device(tokenizer(language, max_length=77, padding='max_length',
truncation=True, return_tensors='pt'), device)
with torch.no_grad():
embeddings = model(inputs)
print("Video x Text: \n",
torch.softmax(embeddings['video'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Image x Text: \n",
torch.softmax(embeddings['image'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Depth x Text: \n",
torch.softmax(embeddings['depth'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Audio x Text: \n",
torch.softmax(embeddings['audio'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Thermal x Text: \n",
torch.softmax(embeddings['thermal'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
```
Then returns the following result.
```bash
Video x Text:
[[9.9989331e-01 1.0667283e-04]
[1.3255903e-03 9.9867439e-01]]
Image x Text:
[[9.9990666e-01 9.3292067e-05]
[4.6132666e-08 1.0000000e+00]]
Depth x Text:
[[0.9954276 0.00457235]
[0.12042473 0.8795753 ]]
Audio x Text:
[[0.97634876 0.02365119]
[0.02917843 0.97082156]]
Thermal x Text:
[[0.9482511 0.0517489 ]
[0.48746133 0.5125386 ]]
```
### Emergency zero-shot
Since languagebind binds each modality together, we also found the **emergency zero-shot**. It's very simple to use.
```python
print("Video x Audio: \n", torch.softmax(embeddings['video'] @ embeddings['audio'].T, dim=-1).detach().cpu().numpy())
print("Image x Depth: \n", torch.softmax(embeddings['image'] @ embeddings['depth'].T, dim=-1).detach().cpu().numpy())
print("Image x Thermal: \n", torch.softmax(embeddings['image'] @ embeddings['thermal'].T, dim=-1).detach().cpu().numpy())
```
Then, you will get:
```
Video x Audio:
[[1.0000000e+00 0.0000000e+00]
[3.1150486e-32 1.0000000e+00]]
Image x Depth:
[[1. 0.]
[0. 1.]]
Image x Thermal:
[[1. 0.]
[0. 1.]]
```
### Different branches for X-Language task
Additionally, LanguageBind can be **disassembled into different branches** to handle different tasks. Note that we do not train Image, which just initialize from OpenCLIP.
#### Thermal
```python
import torch
from languagebind import LanguageBindThermal, LanguageBindThermalTokenizer, LanguageBindThermalProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Thermal'
model = LanguageBindThermal.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindThermalTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
thermal_process = LanguageBindThermalProcessor(model.config, tokenizer)
model.eval()
data = thermal_process([r"your/thermal.jpg"], ['your text'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Depth
```python
import torch
from languagebind import LanguageBindDepth, LanguageBindDepthTokenizer, LanguageBindDepthProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Depth'
model = LanguageBindDepth.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindDepthTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
depth_process = LanguageBindDepthProcessor(model.config, tokenizer)
model.eval()
data = depth_process([r"your/depth.png"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Video
```python
import torch
from languagebind import LanguageBindVideo, LanguageBindVideoTokenizer, LanguageBindVideoProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Video_FT' # also 'LanguageBind/LanguageBind_Video'
model = LanguageBindVideo.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindVideoTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
video_process = LanguageBindVideoProcessor(model.config, tokenizer)
model.eval()
data = video_process(["your/video.mp4"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Audio
```python
import torch
from languagebind import LanguageBindAudio, LanguageBindAudioTokenizer, LanguageBindAudioProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Audio_FT' # also 'LanguageBind/LanguageBind_Audio'
model = LanguageBindAudio.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindAudioTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
audio_process = LanguageBindAudioProcessor(model.config, tokenizer)
model.eval()
data = audio_process([r"your/audio.wav"], ['your audio.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Image
Note that our image encoder is the same as OpenCLIP. **Not** as fine-tuned as other modalities.
```python
import torch
from languagebind import LanguageBindImage, LanguageBindImageTokenizer, LanguageBindImageProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Image'
model = LanguageBindImage.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
image_process = LanguageBindImageProcessor(model.config, tokenizer)
model.eval()
data = image_process([r"your/image.jpg"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
## ๐ฅ VIDAL-10M
The datasets is in [DATASETS.md](DATASETS.md).
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN_AND_VALIDATE.md](TRAIN_AND_VALIDATE.md).
## ๐ Acknowledgement
* [OpenCLIP](https://github.com/mlfoundations/open_clip) An open source pretraining framework.
* [CLIP4Clip](https://github.com/ArrowLuo/CLIP4Clip) An open source Video-Text retrieval framework.
* [sRGB-TIR](https://github.com/rpmsnu/sRGB-TIR) An open source framework to generate infrared (thermal) images.
* [GLPN](https://github.com/vinvino02/GLPDepth) An open source framework to generate depth images.
## ๐ License
* The majority of this project is released under the MIT license as found in the [LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/LICENSE) file.
* The dataset of this project is released under the CC-BY-NC 4.0 license as found in the [DATASET_LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/DATASET_LICENSE) file.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{zhu2023languagebind,
title={LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment},
author={Bin Zhu and Bin Lin and Munan Ning and Yang Yan and Jiaxi Cui and Wang HongFa and Yatian Pang and Wenhao Jiang and Junwu Zhang and Zongwei Li and Cai Wan Zhang and Zhifeng Li and Wei Liu and Li Yuan},
year={2023},
eprint={2310.01852},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/LanguageBind&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/LanguageBind/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/LanguageBind" />
</a>
|
LanguageBind/LanguageBind_Depth | LanguageBind | 2024-02-01T06:57:09Z | 244 | 0 | transformers | [
"transformers",
"pytorch",
"LanguageBindDepth",
"zero-shot-image-classification",
"arxiv:2310.01852",
"license:mit",
"endpoints_compatible",
"region:us"
]
| zero-shot-image-classification | 2023-10-06T09:07:38Z | ---
license: mit
---
<p align="center">
<img src="https://s11.ax1x.com/2024/02/01/pFMDAm9.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/pdf/2310.01852.pdf">ใICLR 2024 ๐ฅใLanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
## ๐ฐ News
* **[2024.01.27]** ๐๐๐ Our [MoE-LLaVA](https://github.com/PKU-YuanGroup/MoE-LLaVA) is released! A sparse model with 3B parameters outperformed the dense model with 7B parameters.
* **[2024.01.16]** ๐ฅ๐ฅ๐ฅ Our LanguageBind has been accepted at ICLR 2024! We earn the score of 6(3)8(6)6(6)6(6) [here](https://openreview.net/forum?id=QmZKc7UZCy¬eId=OgsxQxAleA).
* **[2023.12.15]** ๐ช๐ช๐ช We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M video-text data**. We launch **LanguageBind_Video 1.5**, checking our [model zoo](#-model-zoo).
* **[2023.12.10]** We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M depth and 10M thermal data**. We are in the process of uploading thermal and depth data on [Hugging Face](https://huggingface.co/datasets/LanguageBind/VIDAL-Depth-Thermal) and expect the whole process to last 1-2 months.
* **[2023.11.27]** ๐ฅ๐ฅ๐ฅ We have updated our [paper](https://arxiv.org/abs/2310.01852) with emergency zero-shot results., checking our โจ [results](#emergency-results).
* **[2023.11.26]** ๐ฅ๐ฅ๐ฅ We have open-sourced all textual sources and corresponding YouTube IDs [here](DATASETS.md).
* **[2023.11.26]** ๐ฃ๐ฃ๐ฃ We have open-sourced fully fine-tuned **Video & Audio**, achieving improved performance once again, checking our [model zoo](#-model-zoo).
* **[2023.11.22]** We are about to release a fully fine-tuned version, and the **HUGE** version is currently undergoing training.
* **[2023.11.21]** ๐ฅ We are releasing sample data in [DATASETS.md](DATASETS.md) so that individuals who are interested can further modify the code to train it on their own data.
* **[2023.11.20]** ๐๐๐ [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) builds a large visual-language model to achieve ๐SOTA performances based on LanguageBind encoders.
* **[2023.10.23]** ๐ถ LanguageBind-Audio achieves ๐๐๐**state-of-the-art (SOTA) performance on 5 datasets**, checking our โจ [results](#multiple-modalities)!
* **[2023.10.14]** ๐ฑ Released a stronger LanguageBind-Video, checking our โจ [results](#video-language)! The video checkpoint **have updated** on Huggingface Model Hub!
* **[2023.10.10]** We provide sample data, which can be found in [assets](assets), and [emergency zero-shot usage](#emergency-zero-shot) is described.
* **[2023.10.07]** The checkpoints are available on ๐ค [Huggingface Model](https://huggingface.co/LanguageBind).
* **[2023.10.04]** Code and [demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
### ๐ก High performance, but NO intermediate modality required
LanguageBind is a **language-centric** multimodal pretraining approach, **taking the language as the bind across different modalities** because the language modality is well-explored and contains rich semantics.
* The following first figure shows the architecture of LanguageBind. LanguageBind can be easily extended to segmentation, detection tasks, and potentially to unlimited modalities.
### โก๏ธ A multimodal, fully aligned and voluminous dataset
We propose **VIDAL-10M**, **10 Million data** with **V**ideo, **I**nfrared, **D**epth, **A**udio and their corresponding **L**anguage, which greatly expands the data beyond visual modalities.
* The second figure shows our proposed VIDAL-10M dataset, which includes five modalities: video, infrared, depth, audio, and language.
### ๐ฅ Multi-view enhanced description for training
We make multi-view enhancements to language. We produce multi-view description that combines **meta-data**, **spatial**, and **temporal** to greatly enhance the semantic information of the language. In addition we further **enhance the language with ChatGPT** to create a good semantic space for each modality aligned language.
## ๐ค Demo
* **Local demo.** Highly recommend trying out our web demo, which incorporates all features currently supported by LanguageBind.
```bash
python gradio_app.py
```
* **Online demo.** We provide the [online demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) in Huggingface Spaces. In this demo, you can calculate the similarity of modalities to language, such as audio-to-language, video-to-language, and depth-to-image.
## ๐ ๏ธ Requirements and Installation
* Python >= 3.8
* Pytorch >= 1.13.1
* CUDA Version >= 11.6
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/LanguageBind
cd LanguageBind
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116
pip install -r requirements.txt
```
## ๐ณ Model Zoo
The names in the table represent different encoder models. For example, `LanguageBind/LanguageBind_Video_FT` represents the fully fine-tuned version, while `LanguageBind/LanguageBind_Video` represents the LoRA-tuned version.
You can freely replace them in the recommended [API usage](#-api). We recommend using the fully fine-tuned version, as it offers stronger performance.
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Modality</th><th>LoRA tuning</th><th>Fine-tuning</th>
</tr>
<tr align="center">
<td>Video</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">LanguageBind_Video</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">LanguageBind_Video_FT</a></td>
</tr>
<tr align="center">
<td>Audio</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio">LanguageBind_Audio</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio_FT">LanguageBind_Audio_FT</a></td>
</tr>
<tr align="center">
<td>Depth</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Depth">LanguageBind_Depth</a></td><td>-</td>
</tr>
<tr align="center">
<td>Thermal</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Thermal">LanguageBind_Thermal</a></td><td>-</td>
</tr>
</table>
</div>
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Version</th><th>Tuning</th><th>Model size</th><th>Num_frames</th><th>HF Link</th><th>MSR-VTT</th><th>DiDeMo</th><th>ActivityNet</th><th>MSVD</th>
</tr>
<tr align="center">
<td>LanguageBind_Video</td><td>LoRA</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">Link</a></td><td>42.6</td><td>37.8</td><td>35.1</td><td>52.2</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">Link</a></td><td>42.7</td><td>38.1</td><td>36.9</td><td>53.5</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_V1.5_FT">Link</a></td><td>42.8</td><td>39.7</td><td>38.4</td><td>54.1</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>12</td><td>Coming soon</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_Huge_V1.5_FT">Link</a></td><td>44.8</td><td>39.9</td><td>41.0</td><td>53.7</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>12</td><td>Coming soon</td>
</tr>
</table>
</div>
## ๐ค API
**We open source all modalities preprocessing code.** If you want to load the model (e.g. ```LanguageBind/LanguageBind_Thermal```) from the model hub on Huggingface or on local, you can use the following code snippets!
### Inference for Multi-modal Binding
We have provided some sample datasets in [assets](assets) to quickly see how languagebind works.
```python
import torch
from languagebind import LanguageBind, to_device, transform_dict, LanguageBindImageTokenizer
if __name__ == '__main__':
device = 'cuda:0'
device = torch.device(device)
clip_type = {
'video': 'LanguageBind_Video_FT', # also LanguageBind_Video
'audio': 'LanguageBind_Audio_FT', # also LanguageBind_Audio
'thermal': 'LanguageBind_Thermal',
'image': 'LanguageBind_Image',
'depth': 'LanguageBind_Depth',
}
model = LanguageBind(clip_type=clip_type, cache_dir='./cache_dir')
model = model.to(device)
model.eval()
pretrained_ckpt = f'lb203/LanguageBind_Image'
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir/tokenizer_cache_dir')
modality_transform = {c: transform_dict[c](model.modality_config[c]) for c in clip_type.keys()}
image = ['assets/image/0.jpg', 'assets/image/1.jpg']
audio = ['assets/audio/0.wav', 'assets/audio/1.wav']
video = ['assets/video/0.mp4', 'assets/video/1.mp4']
depth = ['assets/depth/0.png', 'assets/depth/1.png']
thermal = ['assets/thermal/0.jpg', 'assets/thermal/1.jpg']
language = ["Training a parakeet to climb up a ladder.", 'A lion climbing a tree to catch a monkey.']
inputs = {
'image': to_device(modality_transform['image'](image), device),
'video': to_device(modality_transform['video'](video), device),
'audio': to_device(modality_transform['audio'](audio), device),
'depth': to_device(modality_transform['depth'](depth), device),
'thermal': to_device(modality_transform['thermal'](thermal), device),
}
inputs['language'] = to_device(tokenizer(language, max_length=77, padding='max_length',
truncation=True, return_tensors='pt'), device)
with torch.no_grad():
embeddings = model(inputs)
print("Video x Text: \n",
torch.softmax(embeddings['video'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Image x Text: \n",
torch.softmax(embeddings['image'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Depth x Text: \n",
torch.softmax(embeddings['depth'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Audio x Text: \n",
torch.softmax(embeddings['audio'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Thermal x Text: \n",
torch.softmax(embeddings['thermal'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
```
Then returns the following result.
```bash
Video x Text:
[[9.9989331e-01 1.0667283e-04]
[1.3255903e-03 9.9867439e-01]]
Image x Text:
[[9.9990666e-01 9.3292067e-05]
[4.6132666e-08 1.0000000e+00]]
Depth x Text:
[[0.9954276 0.00457235]
[0.12042473 0.8795753 ]]
Audio x Text:
[[0.97634876 0.02365119]
[0.02917843 0.97082156]]
Thermal x Text:
[[0.9482511 0.0517489 ]
[0.48746133 0.5125386 ]]
```
### Emergency zero-shot
Since languagebind binds each modality together, we also found the **emergency zero-shot**. It's very simple to use.
```python
print("Video x Audio: \n", torch.softmax(embeddings['video'] @ embeddings['audio'].T, dim=-1).detach().cpu().numpy())
print("Image x Depth: \n", torch.softmax(embeddings['image'] @ embeddings['depth'].T, dim=-1).detach().cpu().numpy())
print("Image x Thermal: \n", torch.softmax(embeddings['image'] @ embeddings['thermal'].T, dim=-1).detach().cpu().numpy())
```
Then, you will get:
```
Video x Audio:
[[1.0000000e+00 0.0000000e+00]
[3.1150486e-32 1.0000000e+00]]
Image x Depth:
[[1. 0.]
[0. 1.]]
Image x Thermal:
[[1. 0.]
[0. 1.]]
```
### Different branches for X-Language task
Additionally, LanguageBind can be **disassembled into different branches** to handle different tasks. Note that we do not train Image, which just initialize from OpenCLIP.
#### Thermal
```python
import torch
from languagebind import LanguageBindThermal, LanguageBindThermalTokenizer, LanguageBindThermalProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Thermal'
model = LanguageBindThermal.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindThermalTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
thermal_process = LanguageBindThermalProcessor(model.config, tokenizer)
model.eval()
data = thermal_process([r"your/thermal.jpg"], ['your text'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Depth
```python
import torch
from languagebind import LanguageBindDepth, LanguageBindDepthTokenizer, LanguageBindDepthProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Depth'
model = LanguageBindDepth.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindDepthTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
depth_process = LanguageBindDepthProcessor(model.config, tokenizer)
model.eval()
data = depth_process([r"your/depth.png"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Video
```python
import torch
from languagebind import LanguageBindVideo, LanguageBindVideoTokenizer, LanguageBindVideoProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Video_FT' # also 'LanguageBind/LanguageBind_Video'
model = LanguageBindVideo.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindVideoTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
video_process = LanguageBindVideoProcessor(model.config, tokenizer)
model.eval()
data = video_process(["your/video.mp4"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Audio
```python
import torch
from languagebind import LanguageBindAudio, LanguageBindAudioTokenizer, LanguageBindAudioProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Audio_FT' # also 'LanguageBind/LanguageBind_Audio'
model = LanguageBindAudio.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindAudioTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
audio_process = LanguageBindAudioProcessor(model.config, tokenizer)
model.eval()
data = audio_process([r"your/audio.wav"], ['your audio.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Image
Note that our image encoder is the same as OpenCLIP. **Not** as fine-tuned as other modalities.
```python
import torch
from languagebind import LanguageBindImage, LanguageBindImageTokenizer, LanguageBindImageProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Image'
model = LanguageBindImage.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
image_process = LanguageBindImageProcessor(model.config, tokenizer)
model.eval()
data = image_process([r"your/image.jpg"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
## ๐ฅ VIDAL-10M
The datasets is in [DATASETS.md](DATASETS.md).
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN_AND_VALIDATE.md](TRAIN_AND_VALIDATE.md).
## ๐ Acknowledgement
* [OpenCLIP](https://github.com/mlfoundations/open_clip) An open source pretraining framework.
* [CLIP4Clip](https://github.com/ArrowLuo/CLIP4Clip) An open source Video-Text retrieval framework.
* [sRGB-TIR](https://github.com/rpmsnu/sRGB-TIR) An open source framework to generate infrared (thermal) images.
* [GLPN](https://github.com/vinvino02/GLPDepth) An open source framework to generate depth images.
## ๐ License
* The majority of this project is released under the MIT license as found in the [LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/LICENSE) file.
* The dataset of this project is released under the CC-BY-NC 4.0 license as found in the [DATASET_LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/DATASET_LICENSE) file.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{zhu2023languagebind,
title={LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment},
author={Bin Zhu and Bin Lin and Munan Ning and Yang Yan and Jiaxi Cui and Wang HongFa and Yatian Pang and Wenhao Jiang and Junwu Zhang and Zongwei Li and Cai Wan Zhang and Zhifeng Li and Wei Liu and Li Yuan},
year={2023},
eprint={2310.01852},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/LanguageBind&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/LanguageBind/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/LanguageBind" />
</a>
|
LanguageBind/LanguageBind_Audio | LanguageBind | 2024-02-01T06:56:55Z | 549 | 3 | transformers | [
"transformers",
"pytorch",
"LanguageBindAudio",
"zero-shot-image-classification",
"arxiv:2310.01852",
"license:mit",
"endpoints_compatible",
"region:us"
]
| zero-shot-image-classification | 2023-10-06T09:05:22Z | ---
license: mit
---
<p align="center">
<img src="https://s11.ax1x.com/2024/02/01/pFMDAm9.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/pdf/2310.01852.pdf">ใICLR 2024 ๐ฅใLanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
## ๐ฐ News
* **[2024.01.27]** ๐๐๐ Our [MoE-LLaVA](https://github.com/PKU-YuanGroup/MoE-LLaVA) is released! A sparse model with 3B parameters outperformed the dense model with 7B parameters.
* **[2024.01.16]** ๐ฅ๐ฅ๐ฅ Our LanguageBind has been accepted at ICLR 2024! We earn the score of 6(3)8(6)6(6)6(6) [here](https://openreview.net/forum?id=QmZKc7UZCy¬eId=OgsxQxAleA).
* **[2023.12.15]** ๐ช๐ช๐ช We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M video-text data**. We launch **LanguageBind_Video 1.5**, checking our [model zoo](#-model-zoo).
* **[2023.12.10]** We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M depth and 10M thermal data**. We are in the process of uploading thermal and depth data on [Hugging Face](https://huggingface.co/datasets/LanguageBind/VIDAL-Depth-Thermal) and expect the whole process to last 1-2 months.
* **[2023.11.27]** ๐ฅ๐ฅ๐ฅ We have updated our [paper](https://arxiv.org/abs/2310.01852) with emergency zero-shot results., checking our โจ [results](#emergency-results).
* **[2023.11.26]** ๐ฅ๐ฅ๐ฅ We have open-sourced all textual sources and corresponding YouTube IDs [here](DATASETS.md).
* **[2023.11.26]** ๐ฃ๐ฃ๐ฃ We have open-sourced fully fine-tuned **Video & Audio**, achieving improved performance once again, checking our [model zoo](#-model-zoo).
* **[2023.11.22]** We are about to release a fully fine-tuned version, and the **HUGE** version is currently undergoing training.
* **[2023.11.21]** ๐ฅ We are releasing sample data in [DATASETS.md](DATASETS.md) so that individuals who are interested can further modify the code to train it on their own data.
* **[2023.11.20]** ๐๐๐ [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) builds a large visual-language model to achieve ๐SOTA performances based on LanguageBind encoders.
* **[2023.10.23]** ๐ถ LanguageBind-Audio achieves ๐๐๐**state-of-the-art (SOTA) performance on 5 datasets**, checking our โจ [results](#multiple-modalities)!
* **[2023.10.14]** ๐ฑ Released a stronger LanguageBind-Video, checking our โจ [results](#video-language)! The video checkpoint **have updated** on Huggingface Model Hub!
* **[2023.10.10]** We provide sample data, which can be found in [assets](assets), and [emergency zero-shot usage](#emergency-zero-shot) is described.
* **[2023.10.07]** The checkpoints are available on ๐ค [Huggingface Model](https://huggingface.co/LanguageBind).
* **[2023.10.04]** Code and [demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
### ๐ก High performance, but NO intermediate modality required
LanguageBind is a **language-centric** multimodal pretraining approach, **taking the language as the bind across different modalities** because the language modality is well-explored and contains rich semantics.
* The following first figure shows the architecture of LanguageBind. LanguageBind can be easily extended to segmentation, detection tasks, and potentially to unlimited modalities.
### โก๏ธ A multimodal, fully aligned and voluminous dataset
We propose **VIDAL-10M**, **10 Million data** with **V**ideo, **I**nfrared, **D**epth, **A**udio and their corresponding **L**anguage, which greatly expands the data beyond visual modalities.
* The second figure shows our proposed VIDAL-10M dataset, which includes five modalities: video, infrared, depth, audio, and language.
### ๐ฅ Multi-view enhanced description for training
We make multi-view enhancements to language. We produce multi-view description that combines **meta-data**, **spatial**, and **temporal** to greatly enhance the semantic information of the language. In addition we further **enhance the language with ChatGPT** to create a good semantic space for each modality aligned language.
## ๐ค Demo
* **Local demo.** Highly recommend trying out our web demo, which incorporates all features currently supported by LanguageBind.
```bash
python gradio_app.py
```
* **Online demo.** We provide the [online demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) in Huggingface Spaces. In this demo, you can calculate the similarity of modalities to language, such as audio-to-language, video-to-language, and depth-to-image.
## ๐ ๏ธ Requirements and Installation
* Python >= 3.8
* Pytorch >= 1.13.1
* CUDA Version >= 11.6
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/LanguageBind
cd LanguageBind
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116
pip install -r requirements.txt
```
## ๐ณ Model Zoo
The names in the table represent different encoder models. For example, `LanguageBind/LanguageBind_Video_FT` represents the fully fine-tuned version, while `LanguageBind/LanguageBind_Video` represents the LoRA-tuned version.
You can freely replace them in the recommended [API usage](#-api). We recommend using the fully fine-tuned version, as it offers stronger performance.
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Modality</th><th>LoRA tuning</th><th>Fine-tuning</th>
</tr>
<tr align="center">
<td>Video</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">LanguageBind_Video</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">LanguageBind_Video_FT</a></td>
</tr>
<tr align="center">
<td>Audio</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio">LanguageBind_Audio</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio_FT">LanguageBind_Audio_FT</a></td>
</tr>
<tr align="center">
<td>Depth</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Depth">LanguageBind_Depth</a></td><td>-</td>
</tr>
<tr align="center">
<td>Thermal</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Thermal">LanguageBind_Thermal</a></td><td>-</td>
</tr>
</table>
</div>
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Version</th><th>Tuning</th><th>Model size</th><th>Num_frames</th><th>HF Link</th><th>MSR-VTT</th><th>DiDeMo</th><th>ActivityNet</th><th>MSVD</th>
</tr>
<tr align="center">
<td>LanguageBind_Video</td><td>LoRA</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">Link</a></td><td>42.6</td><td>37.8</td><td>35.1</td><td>52.2</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">Link</a></td><td>42.7</td><td>38.1</td><td>36.9</td><td>53.5</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_V1.5_FT">Link</a></td><td>42.8</td><td>39.7</td><td>38.4</td><td>54.1</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>12</td><td>Coming soon</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_Huge_V1.5_FT">Link</a></td><td>44.8</td><td>39.9</td><td>41.0</td><td>53.7</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>12</td><td>Coming soon</td>
</tr>
</table>
</div>
## ๐ค API
**We open source all modalities preprocessing code.** If you want to load the model (e.g. ```LanguageBind/LanguageBind_Thermal```) from the model hub on Huggingface or on local, you can use the following code snippets!
### Inference for Multi-modal Binding
We have provided some sample datasets in [assets](assets) to quickly see how languagebind works.
```python
import torch
from languagebind import LanguageBind, to_device, transform_dict, LanguageBindImageTokenizer
if __name__ == '__main__':
device = 'cuda:0'
device = torch.device(device)
clip_type = {
'video': 'LanguageBind_Video_FT', # also LanguageBind_Video
'audio': 'LanguageBind_Audio_FT', # also LanguageBind_Audio
'thermal': 'LanguageBind_Thermal',
'image': 'LanguageBind_Image',
'depth': 'LanguageBind_Depth',
}
model = LanguageBind(clip_type=clip_type, cache_dir='./cache_dir')
model = model.to(device)
model.eval()
pretrained_ckpt = f'lb203/LanguageBind_Image'
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir/tokenizer_cache_dir')
modality_transform = {c: transform_dict[c](model.modality_config[c]) for c in clip_type.keys()}
image = ['assets/image/0.jpg', 'assets/image/1.jpg']
audio = ['assets/audio/0.wav', 'assets/audio/1.wav']
video = ['assets/video/0.mp4', 'assets/video/1.mp4']
depth = ['assets/depth/0.png', 'assets/depth/1.png']
thermal = ['assets/thermal/0.jpg', 'assets/thermal/1.jpg']
language = ["Training a parakeet to climb up a ladder.", 'A lion climbing a tree to catch a monkey.']
inputs = {
'image': to_device(modality_transform['image'](image), device),
'video': to_device(modality_transform['video'](video), device),
'audio': to_device(modality_transform['audio'](audio), device),
'depth': to_device(modality_transform['depth'](depth), device),
'thermal': to_device(modality_transform['thermal'](thermal), device),
}
inputs['language'] = to_device(tokenizer(language, max_length=77, padding='max_length',
truncation=True, return_tensors='pt'), device)
with torch.no_grad():
embeddings = model(inputs)
print("Video x Text: \n",
torch.softmax(embeddings['video'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Image x Text: \n",
torch.softmax(embeddings['image'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Depth x Text: \n",
torch.softmax(embeddings['depth'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Audio x Text: \n",
torch.softmax(embeddings['audio'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Thermal x Text: \n",
torch.softmax(embeddings['thermal'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
```
Then returns the following result.
```bash
Video x Text:
[[9.9989331e-01 1.0667283e-04]
[1.3255903e-03 9.9867439e-01]]
Image x Text:
[[9.9990666e-01 9.3292067e-05]
[4.6132666e-08 1.0000000e+00]]
Depth x Text:
[[0.9954276 0.00457235]
[0.12042473 0.8795753 ]]
Audio x Text:
[[0.97634876 0.02365119]
[0.02917843 0.97082156]]
Thermal x Text:
[[0.9482511 0.0517489 ]
[0.48746133 0.5125386 ]]
```
### Emergency zero-shot
Since languagebind binds each modality together, we also found the **emergency zero-shot**. It's very simple to use.
```python
print("Video x Audio: \n", torch.softmax(embeddings['video'] @ embeddings['audio'].T, dim=-1).detach().cpu().numpy())
print("Image x Depth: \n", torch.softmax(embeddings['image'] @ embeddings['depth'].T, dim=-1).detach().cpu().numpy())
print("Image x Thermal: \n", torch.softmax(embeddings['image'] @ embeddings['thermal'].T, dim=-1).detach().cpu().numpy())
```
Then, you will get:
```
Video x Audio:
[[1.0000000e+00 0.0000000e+00]
[3.1150486e-32 1.0000000e+00]]
Image x Depth:
[[1. 0.]
[0. 1.]]
Image x Thermal:
[[1. 0.]
[0. 1.]]
```
### Different branches for X-Language task
Additionally, LanguageBind can be **disassembled into different branches** to handle different tasks. Note that we do not train Image, which just initialize from OpenCLIP.
#### Thermal
```python
import torch
from languagebind import LanguageBindThermal, LanguageBindThermalTokenizer, LanguageBindThermalProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Thermal'
model = LanguageBindThermal.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindThermalTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
thermal_process = LanguageBindThermalProcessor(model.config, tokenizer)
model.eval()
data = thermal_process([r"your/thermal.jpg"], ['your text'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Depth
```python
import torch
from languagebind import LanguageBindDepth, LanguageBindDepthTokenizer, LanguageBindDepthProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Depth'
model = LanguageBindDepth.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindDepthTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
depth_process = LanguageBindDepthProcessor(model.config, tokenizer)
model.eval()
data = depth_process([r"your/depth.png"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Video
```python
import torch
from languagebind import LanguageBindVideo, LanguageBindVideoTokenizer, LanguageBindVideoProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Video_FT' # also 'LanguageBind/LanguageBind_Video'
model = LanguageBindVideo.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindVideoTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
video_process = LanguageBindVideoProcessor(model.config, tokenizer)
model.eval()
data = video_process(["your/video.mp4"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Audio
```python
import torch
from languagebind import LanguageBindAudio, LanguageBindAudioTokenizer, LanguageBindAudioProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Audio_FT' # also 'LanguageBind/LanguageBind_Audio'
model = LanguageBindAudio.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindAudioTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
audio_process = LanguageBindAudioProcessor(model.config, tokenizer)
model.eval()
data = audio_process([r"your/audio.wav"], ['your audio.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Image
Note that our image encoder is the same as OpenCLIP. **Not** as fine-tuned as other modalities.
```python
import torch
from languagebind import LanguageBindImage, LanguageBindImageTokenizer, LanguageBindImageProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Image'
model = LanguageBindImage.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
image_process = LanguageBindImageProcessor(model.config, tokenizer)
model.eval()
data = image_process([r"your/image.jpg"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
## ๐ฅ VIDAL-10M
The datasets is in [DATASETS.md](DATASETS.md).
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN_AND_VALIDATE.md](TRAIN_AND_VALIDATE.md).
## ๐ Acknowledgement
* [OpenCLIP](https://github.com/mlfoundations/open_clip) An open source pretraining framework.
* [CLIP4Clip](https://github.com/ArrowLuo/CLIP4Clip) An open source Video-Text retrieval framework.
* [sRGB-TIR](https://github.com/rpmsnu/sRGB-TIR) An open source framework to generate infrared (thermal) images.
* [GLPN](https://github.com/vinvino02/GLPDepth) An open source framework to generate depth images.
## ๐ License
* The majority of this project is released under the MIT license as found in the [LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/LICENSE) file.
* The dataset of this project is released under the CC-BY-NC 4.0 license as found in the [DATASET_LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/DATASET_LICENSE) file.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{zhu2023languagebind,
title={LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment},
author={Bin Zhu and Bin Lin and Munan Ning and Yang Yan and Jiaxi Cui and Wang HongFa and Yatian Pang and Wenhao Jiang and Junwu Zhang and Zongwei Li and Cai Wan Zhang and Zhifeng Li and Wei Liu and Li Yuan},
year={2023},
eprint={2310.01852},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/LanguageBind&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/LanguageBind/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/LanguageBind" />
</a>
|
LanguageBind/LanguageBind_Audio_FT | LanguageBind | 2024-02-01T06:56:41Z | 5,296 | 1 | transformers | [
"transformers",
"pytorch",
"LanguageBindAudio",
"zero-shot-image-classification",
"arxiv:2310.01852",
"license:mit",
"endpoints_compatible",
"region:us"
]
| zero-shot-image-classification | 2023-11-26T07:37:41Z | ---
license: mit
---
<p align="center">
<img src="https://s11.ax1x.com/2024/02/01/pFMDAm9.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/pdf/2310.01852.pdf">ใICLR 2024 ๐ฅใLanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
## ๐ฐ News
* **[2024.01.27]** ๐๐๐ Our [MoE-LLaVA](https://github.com/PKU-YuanGroup/MoE-LLaVA) is released! A sparse model with 3B parameters outperformed the dense model with 7B parameters.
* **[2024.01.16]** ๐ฅ๐ฅ๐ฅ Our LanguageBind has been accepted at ICLR 2024! We earn the score of 6(3)8(6)6(6)6(6) [here](https://openreview.net/forum?id=QmZKc7UZCy¬eId=OgsxQxAleA).
* **[2023.12.15]** ๐ช๐ช๐ช We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M video-text data**. We launch **LanguageBind_Video 1.5**, checking our [model zoo](#-model-zoo).
* **[2023.12.10]** We expand the ๐ฅ๐ฅ๐ฅ VIDAL dataset and now have **10M depth and 10M thermal data**. We are in the process of uploading thermal and depth data on [Hugging Face](https://huggingface.co/datasets/LanguageBind/VIDAL-Depth-Thermal) and expect the whole process to last 1-2 months.
* **[2023.11.27]** ๐ฅ๐ฅ๐ฅ We have updated our [paper](https://arxiv.org/abs/2310.01852) with emergency zero-shot results., checking our โจ [results](#emergency-results).
* **[2023.11.26]** ๐ฅ๐ฅ๐ฅ We have open-sourced all textual sources and corresponding YouTube IDs [here](DATASETS.md).
* **[2023.11.26]** ๐ฃ๐ฃ๐ฃ We have open-sourced fully fine-tuned **Video & Audio**, achieving improved performance once again, checking our [model zoo](#-model-zoo).
* **[2023.11.22]** We are about to release a fully fine-tuned version, and the **HUGE** version is currently undergoing training.
* **[2023.11.21]** ๐ฅ We are releasing sample data in [DATASETS.md](DATASETS.md) so that individuals who are interested can further modify the code to train it on their own data.
* **[2023.11.20]** ๐๐๐ [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) builds a large visual-language model to achieve ๐SOTA performances based on LanguageBind encoders.
* **[2023.10.23]** ๐ถ LanguageBind-Audio achieves ๐๐๐**state-of-the-art (SOTA) performance on 5 datasets**, checking our โจ [results](#multiple-modalities)!
* **[2023.10.14]** ๐ฑ Released a stronger LanguageBind-Video, checking our โจ [results](#video-language)! The video checkpoint **have updated** on Huggingface Model Hub!
* **[2023.10.10]** We provide sample data, which can be found in [assets](assets), and [emergency zero-shot usage](#emergency-zero-shot) is described.
* **[2023.10.07]** The checkpoints are available on ๐ค [Huggingface Model](https://huggingface.co/LanguageBind).
* **[2023.10.04]** Code and [demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
### ๐ก High performance, but NO intermediate modality required
LanguageBind is a **language-centric** multimodal pretraining approach, **taking the language as the bind across different modalities** because the language modality is well-explored and contains rich semantics.
* The following first figure shows the architecture of LanguageBind. LanguageBind can be easily extended to segmentation, detection tasks, and potentially to unlimited modalities.
### โก๏ธ A multimodal, fully aligned and voluminous dataset
We propose **VIDAL-10M**, **10 Million data** with **V**ideo, **I**nfrared, **D**epth, **A**udio and their corresponding **L**anguage, which greatly expands the data beyond visual modalities.
* The second figure shows our proposed VIDAL-10M dataset, which includes five modalities: video, infrared, depth, audio, and language.
### ๐ฅ Multi-view enhanced description for training
We make multi-view enhancements to language. We produce multi-view description that combines **meta-data**, **spatial**, and **temporal** to greatly enhance the semantic information of the language. In addition we further **enhance the language with ChatGPT** to create a good semantic space for each modality aligned language.
## ๐ค Demo
* **Local demo.** Highly recommend trying out our web demo, which incorporates all features currently supported by LanguageBind.
```bash
python gradio_app.py
```
* **Online demo.** We provide the [online demo](https://huggingface.co/spaces/LanguageBind/LanguageBind) in Huggingface Spaces. In this demo, you can calculate the similarity of modalities to language, such as audio-to-language, video-to-language, and depth-to-image.
## ๐ ๏ธ Requirements and Installation
* Python >= 3.8
* Pytorch >= 1.13.1
* CUDA Version >= 11.6
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/LanguageBind
cd LanguageBind
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116
pip install -r requirements.txt
```
## ๐ณ Model Zoo
The names in the table represent different encoder models. For example, `LanguageBind/LanguageBind_Video_FT` represents the fully fine-tuned version, while `LanguageBind/LanguageBind_Video` represents the LoRA-tuned version.
You can freely replace them in the recommended [API usage](#-api). We recommend using the fully fine-tuned version, as it offers stronger performance.
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Modality</th><th>LoRA tuning</th><th>Fine-tuning</th>
</tr>
<tr align="center">
<td>Video</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">LanguageBind_Video</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">LanguageBind_Video_FT</a></td>
</tr>
<tr align="center">
<td>Audio</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio">LanguageBind_Audio</a></td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Audio_FT">LanguageBind_Audio_FT</a></td>
</tr>
<tr align="center">
<td>Depth</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Depth">LanguageBind_Depth</a></td><td>-</td>
</tr>
<tr align="center">
<td>Thermal</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Thermal">LanguageBind_Thermal</a></td><td>-</td>
</tr>
</table>
</div>
<div align="center">
<table border="1" width="100%">
<tr align="center">
<th>Version</th><th>Tuning</th><th>Model size</th><th>Num_frames</th><th>HF Link</th><th>MSR-VTT</th><th>DiDeMo</th><th>ActivityNet</th><th>MSVD</th>
</tr>
<tr align="center">
<td>LanguageBind_Video</td><td>LoRA</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video">Link</a></td><td>42.6</td><td>37.8</td><td>35.1</td><td>52.2</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_FT">Link</a></td><td>42.7</td><td>38.1</td><td>36.9</td><td>53.5</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_V1.5_FT">Link</a></td><td>42.8</td><td>39.7</td><td>38.4</td><td>54.1</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_V1.5_FT</td><td>Full-tuning</td><td>Large</td><td>12</td><td>Coming soon</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>8</td><td><a href="https://huggingface.co/LanguageBind/LanguageBind_Video_Huge_V1.5_FT">Link</a></td><td>44.8</td><td>39.9</td><td>41.0</td><td>53.7</td>
</tr>
<tr align="center">
<td>LanguageBind_Video_Huge_V1.5_FT</td><td>Full-tuning</td><td>Huge</td><td>12</td><td>Coming soon</td>
</tr>
</table>
</div>
## ๐ค API
**We open source all modalities preprocessing code.** If you want to load the model (e.g. ```LanguageBind/LanguageBind_Thermal```) from the model hub on Huggingface or on local, you can use the following code snippets!
### Inference for Multi-modal Binding
We have provided some sample datasets in [assets](assets) to quickly see how languagebind works.
```python
import torch
from languagebind import LanguageBind, to_device, transform_dict, LanguageBindImageTokenizer
if __name__ == '__main__':
device = 'cuda:0'
device = torch.device(device)
clip_type = {
'video': 'LanguageBind_Video_FT', # also LanguageBind_Video
'audio': 'LanguageBind_Audio_FT', # also LanguageBind_Audio
'thermal': 'LanguageBind_Thermal',
'image': 'LanguageBind_Image',
'depth': 'LanguageBind_Depth',
}
model = LanguageBind(clip_type=clip_type, cache_dir='./cache_dir')
model = model.to(device)
model.eval()
pretrained_ckpt = f'lb203/LanguageBind_Image'
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir/tokenizer_cache_dir')
modality_transform = {c: transform_dict[c](model.modality_config[c]) for c in clip_type.keys()}
image = ['assets/image/0.jpg', 'assets/image/1.jpg']
audio = ['assets/audio/0.wav', 'assets/audio/1.wav']
video = ['assets/video/0.mp4', 'assets/video/1.mp4']
depth = ['assets/depth/0.png', 'assets/depth/1.png']
thermal = ['assets/thermal/0.jpg', 'assets/thermal/1.jpg']
language = ["Training a parakeet to climb up a ladder.", 'A lion climbing a tree to catch a monkey.']
inputs = {
'image': to_device(modality_transform['image'](image), device),
'video': to_device(modality_transform['video'](video), device),
'audio': to_device(modality_transform['audio'](audio), device),
'depth': to_device(modality_transform['depth'](depth), device),
'thermal': to_device(modality_transform['thermal'](thermal), device),
}
inputs['language'] = to_device(tokenizer(language, max_length=77, padding='max_length',
truncation=True, return_tensors='pt'), device)
with torch.no_grad():
embeddings = model(inputs)
print("Video x Text: \n",
torch.softmax(embeddings['video'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Image x Text: \n",
torch.softmax(embeddings['image'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Depth x Text: \n",
torch.softmax(embeddings['depth'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Audio x Text: \n",
torch.softmax(embeddings['audio'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
print("Thermal x Text: \n",
torch.softmax(embeddings['thermal'] @ embeddings['language'].T, dim=-1).detach().cpu().numpy())
```
Then returns the following result.
```bash
Video x Text:
[[9.9989331e-01 1.0667283e-04]
[1.3255903e-03 9.9867439e-01]]
Image x Text:
[[9.9990666e-01 9.3292067e-05]
[4.6132666e-08 1.0000000e+00]]
Depth x Text:
[[0.9954276 0.00457235]
[0.12042473 0.8795753 ]]
Audio x Text:
[[0.97634876 0.02365119]
[0.02917843 0.97082156]]
Thermal x Text:
[[0.9482511 0.0517489 ]
[0.48746133 0.5125386 ]]
```
### Emergency zero-shot
Since languagebind binds each modality together, we also found the **emergency zero-shot**. It's very simple to use.
```python
print("Video x Audio: \n", torch.softmax(embeddings['video'] @ embeddings['audio'].T, dim=-1).detach().cpu().numpy())
print("Image x Depth: \n", torch.softmax(embeddings['image'] @ embeddings['depth'].T, dim=-1).detach().cpu().numpy())
print("Image x Thermal: \n", torch.softmax(embeddings['image'] @ embeddings['thermal'].T, dim=-1).detach().cpu().numpy())
```
Then, you will get:
```
Video x Audio:
[[1.0000000e+00 0.0000000e+00]
[3.1150486e-32 1.0000000e+00]]
Image x Depth:
[[1. 0.]
[0. 1.]]
Image x Thermal:
[[1. 0.]
[0. 1.]]
```
### Different branches for X-Language task
Additionally, LanguageBind can be **disassembled into different branches** to handle different tasks. Note that we do not train Image, which just initialize from OpenCLIP.
#### Thermal
```python
import torch
from languagebind import LanguageBindThermal, LanguageBindThermalTokenizer, LanguageBindThermalProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Thermal'
model = LanguageBindThermal.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindThermalTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
thermal_process = LanguageBindThermalProcessor(model.config, tokenizer)
model.eval()
data = thermal_process([r"your/thermal.jpg"], ['your text'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Depth
```python
import torch
from languagebind import LanguageBindDepth, LanguageBindDepthTokenizer, LanguageBindDepthProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Depth'
model = LanguageBindDepth.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindDepthTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
depth_process = LanguageBindDepthProcessor(model.config, tokenizer)
model.eval()
data = depth_process([r"your/depth.png"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Video
```python
import torch
from languagebind import LanguageBindVideo, LanguageBindVideoTokenizer, LanguageBindVideoProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Video_FT' # also 'LanguageBind/LanguageBind_Video'
model = LanguageBindVideo.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindVideoTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
video_process = LanguageBindVideoProcessor(model.config, tokenizer)
model.eval()
data = video_process(["your/video.mp4"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Audio
```python
import torch
from languagebind import LanguageBindAudio, LanguageBindAudioTokenizer, LanguageBindAudioProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Audio_FT' # also 'LanguageBind/LanguageBind_Audio'
model = LanguageBindAudio.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindAudioTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
audio_process = LanguageBindAudioProcessor(model.config, tokenizer)
model.eval()
data = audio_process([r"your/audio.wav"], ['your audio.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
#### Image
Note that our image encoder is the same as OpenCLIP. **Not** as fine-tuned as other modalities.
```python
import torch
from languagebind import LanguageBindImage, LanguageBindImageTokenizer, LanguageBindImageProcessor
pretrained_ckpt = 'LanguageBind/LanguageBind_Image'
model = LanguageBindImage.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
tokenizer = LanguageBindImageTokenizer.from_pretrained(pretrained_ckpt, cache_dir='./cache_dir')
image_process = LanguageBindImageProcessor(model.config, tokenizer)
model.eval()
data = image_process([r"your/image.jpg"], ['your text.'], return_tensors='pt')
with torch.no_grad():
out = model(**data)
print(out.text_embeds @ out.image_embeds.T)
```
## ๐ฅ VIDAL-10M
The datasets is in [DATASETS.md](DATASETS.md).
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN_AND_VALIDATE.md](TRAIN_AND_VALIDATE.md).
## ๐ Acknowledgement
* [OpenCLIP](https://github.com/mlfoundations/open_clip) An open source pretraining framework.
* [CLIP4Clip](https://github.com/ArrowLuo/CLIP4Clip) An open source Video-Text retrieval framework.
* [sRGB-TIR](https://github.com/rpmsnu/sRGB-TIR) An open source framework to generate infrared (thermal) images.
* [GLPN](https://github.com/vinvino02/GLPDepth) An open source framework to generate depth images.
## ๐ License
* The majority of this project is released under the MIT license as found in the [LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/LICENSE) file.
* The dataset of this project is released under the CC-BY-NC 4.0 license as found in the [DATASET_LICENSE](https://github.com/PKU-YuanGroup/LanguageBind/blob/main/DATASET_LICENSE) file.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{zhu2023languagebind,
title={LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment},
author={Bin Zhu and Bin Lin and Munan Ning and Yang Yan and Jiaxi Cui and Wang HongFa and Yatian Pang and Wenhao Jiang and Junwu Zhang and Zongwei Li and Cai Wan Zhang and Zhifeng Li and Wei Liu and Li Yuan},
year={2023},
eprint={2310.01852},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/LanguageBind&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/LanguageBind/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/LanguageBind" />
</a>
|
nakcnx/phi-2-sql-1g10e-lora | nakcnx | 2024-02-01T06:55:46Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| null | 2024-02-01T06:55:25Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
LanguageBind/Video-LLaVA-Pretrain-7B | LanguageBind | 2024-02-01T06:52:24Z | 33 | 10 | transformers | [
"transformers",
"llava",
"text-generation",
"arxiv:2311.10122",
"arxiv:2310.01852",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-generation | 2023-11-17T05:07:08Z | ---
license: apache-2.0
---
<p align="center">
<img src="https://z1.ax1x.com/2023/11/07/pil4sqH.png" width="150" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/abs/2311.10122">Video-LLaVA: Learning United Visual Representation by Alignment Before Projection</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
## ๐ฐ News
* **[2024.01.27]** ๐๐๐ Our [MoE-LLaVA](https://github.com/PKU-YuanGroup/MoE-LLaVA) is released! A sparse model with 3B parameters outperformed the dense model with 7B parameters.
* **[2024.01.17]** ๐ฅ๐ฅ๐ฅ Our [LanguageBind](https://github.com/PKU-YuanGroup/LanguageBind) has been accepted at ICLR 2024!
* **[2024.01.16]** ๐ฅ๐ฅ๐ฅ We reorganize the code and support LoRA fine-tuning, checking [finetune_lora.sh](scripts/v1_5/finetune_lora.sh).
* **[2023.11.30]** ๐ค Thanks to the generous contributions of the community, the [OpenXLab's demo](https://openxlab.org.cn/apps/detail/houshaowei/Video-LLaVA) is now accessible.
* **[2023.11.23]** We are training a new and powerful model.
* **[2023.11.21]** ๐ค Check out the [replicate demo](https://replicate.com/nateraw/video-llava), created by [@nateraw](https://github.com/nateraw), who has generously supported our research!
* **[2023.11.20]** ๐ค [Hugging Face demo](https://huggingface.co/spaces/LanguageBind/Video-LLaVA) and **all codes & datasets** are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
Video-LLaVA exhibits remarkable interactive capabilities between images and videos, despite the absence of image-video pairs in the dataset.
### ๐ก Simple baseline, learning united visual representation by alignment before projection
- With **the binding of unified visual representations to the language feature space**, we enable an LLM to perform visual reasoning capabilities on both images and videos simultaneously.
### ๐ฅ High performance, complementary learning with video and image
- Extensive experiments demonstrate **the complementarity of modalities**, showcasing significant superiority when compared to models specifically designed for either images or videos.
## ๐ค Demo
### Gradio Web UI
Highly recommend trying out our web demo by the following command, which incorporates all features currently supported by Video-LLaVA. We also provide [online demo](https://huggingface.co/spaces/LanguageBind/Video-LLaVA) in Huggingface Spaces.
```bash
python -m videollava.serve.gradio_web_server
```
### CLI Inference
```bash
python -m videollava.serve.cli --model-path "LanguageBind/Video-LLaVA-7B" --file "path/to/your/video.mp4" --load-4bit
```
```bash
python -m videollava.serve.cli --model-path "LanguageBind/Video-LLaVA-7B" --file "path/to/your/image.jpg" --load-4bit
```
## ๐ ๏ธ Requirements and Installation
* Python >= 3.10
* Pytorch == 2.0.1
* CUDA Version >= 11.7
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/Video-LLaVA
cd Video-LLaVA
conda create -n videollava python=3.10 -y
conda activate videollava
pip install --upgrade pip # enable PEP 660 support
pip install -e .
pip install -e ".[train]"
pip install flash-attn --no-build-isolation
pip install decord opencv-python git+https://github.com/facebookresearch/pytorchvideo.git@28fe037d212663c6a24f373b94cc5d478c8c1a1d
```
## ๐ค API
**We open source all codes.** If you want to load the model (e.g. ```LanguageBind/Video-LLaVA-7B```) on local, you can use the following code snippets.
### Inference for image
```python
import torch
from videollava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
from videollava.conversation import conv_templates, SeparatorStyle
from videollava.model.builder import load_pretrained_model
from videollava.utils import disable_torch_init
from videollava.mm_utils import tokenizer_image_token, get_model_name_from_path, KeywordsStoppingCriteria
def main():
disable_torch_init()
image = 'videollava/serve/examples/extreme_ironing.jpg'
inp = 'What is unusual about this image?'
model_path = 'LanguageBind/Video-LLaVA-7B'
cache_dir = 'cache_dir'
device = 'cuda'
load_4bit, load_8bit = True, False
model_name = get_model_name_from_path(model_path)
tokenizer, model, processor, _ = load_pretrained_model(model_path, None, model_name, load_8bit, load_4bit, device=device, cache_dir=cache_dir)
image_processor = processor['image']
conv_mode = "llava_v1"
conv = conv_templates[conv_mode].copy()
roles = conv.roles
image_tensor = image_processor.preprocess(image, return_tensors='pt')['pixel_values']
if type(image_tensor) is list:
tensor = [image.to(model.device, dtype=torch.float16) for image in image_tensor]
else:
tensor = image_tensor.to(model.device, dtype=torch.float16)
print(f"{roles[1]}: {inp}")
inp = DEFAULT_IMAGE_TOKEN + '\n' + inp
conv.append_message(conv.roles[0], inp)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).cuda()
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=tensor,
do_sample=True,
temperature=0.2,
max_new_tokens=1024,
use_cache=True,
stopping_criteria=[stopping_criteria])
outputs = tokenizer.decode(output_ids[0, input_ids.shape[1]:]).strip()
print(outputs)
if __name__ == '__main__':
main()
```
### Inference for video
```python
import torch
from videollava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
from videollava.conversation import conv_templates, SeparatorStyle
from videollava.model.builder import load_pretrained_model
from videollava.utils import disable_torch_init
from videollava.mm_utils import tokenizer_image_token, get_model_name_from_path, KeywordsStoppingCriteria
def main():
disable_torch_init()
video = 'videollava/serve/examples/sample_demo_1.mp4'
inp = 'Why is this video funny?'
model_path = 'LanguageBind/Video-LLaVA-7B'
cache_dir = 'cache_dir'
device = 'cuda'
load_4bit, load_8bit = True, False
model_name = get_model_name_from_path(model_path)
tokenizer, model, processor, _ = load_pretrained_model(model_path, None, model_name, load_8bit, load_4bit, device=device, cache_dir=cache_dir)
video_processor = processor['video']
conv_mode = "llava_v1"
conv = conv_templates[conv_mode].copy()
roles = conv.roles
video_tensor = video_processor(video, return_tensors='pt')['pixel_values']
if type(video_tensor) is list:
tensor = [video.to(model.device, dtype=torch.float16) for video in video_tensor]
else:
tensor = video_tensor.to(model.device, dtype=torch.float16)
print(f"{roles[1]}: {inp}")
inp = ' '.join([DEFAULT_IMAGE_TOKEN] * model.get_video_tower().config.num_frames) + '\n' + inp
conv.append_message(conv.roles[0], inp)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).cuda()
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=tensor,
do_sample=True,
temperature=0.1,
max_new_tokens=1024,
use_cache=True,
stopping_criteria=[stopping_criteria])
outputs = tokenizer.decode(output_ids[0, input_ids.shape[1]:]).strip()
print(outputs)
if __name__ == '__main__':
main()
```
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN_AND_VALIDATE.md](TRAIN_AND_VALIDATE.md).
## ๐ Acknowledgement
* [LLaVA](https://github.com/haotian-liu/LLaVA) The codebase we built upon and it is an efficient large language and vision assistant.
* [Video-ChatGPT](https://github.com/mbzuai-oryx/Video-ChatGPT) Great job contributing the evaluation code and dataset.
## ๐ Related Projects
* [LanguageBind](https://github.com/PKU-YuanGroup/LanguageBind) An open source five modalities language-based retrieval framework.
* [Chat-UniVi](https://github.com/PKU-YuanGroup/Chat-UniVi) This framework empowers the model to efficiently utilize a limited number of visual tokens.
## ๐ License
* The majority of this project is released under the Apache 2.0 license as found in the [LICENSE](https://github.com/PKU-YuanGroup/Video-LLaVA/blob/main/LICENSE) file.
* The service is a research preview intended for non-commercial use only, subject to the model [License](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md) of LLaMA, [Terms of Use](https://openai.com/policies/terms-of-use) of the data generated by OpenAI, and [Privacy Practices](https://chrome.google.com/webstore/detail/sharegpt-share-your-chatg/daiacboceoaocpibfodeljbdfacokfjb) of ShareGPT. Please contact us if you find any potential violation.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@article{lin2023video,
title={Video-LLaVA: Learning United Visual Representation by Alignment Before Projection},
author={Lin, Bin and Zhu, Bin and Ye, Yang and Ning, Munan and Jin, Peng and Yuan, Li},
journal={arXiv preprint arXiv:2311.10122},
year={2023}
}
```
```BibTeX
@article{zhu2023languagebind,
title={LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment},
author={Zhu, Bin and Lin, Bin and Ning, Munan and Yan, Yang and Cui, Jiaxi and Wang, HongFa and Pang, Yatian and Jiang, Wenhao and Zhang, Junwu and Li, Zongwei and others},
journal={arXiv preprint arXiv:2310.01852},
year={2023}
}
```
<!---->
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/Video-LLaVA&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/Video-LLaVA/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/Video-LLaVA" />
</a>
|
JKuang96/pixelcopter_v3 | JKuang96 | 2024-02-01T06:51:02Z | 0 | 0 | null | [
"Pixelcopter-PLE-v0",
"reinforce",
"reinforcement-learning",
"custom-implementation",
"deep-rl-class",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T06:50:56Z | ---
tags:
- Pixelcopter-PLE-v0
- reinforce
- reinforcement-learning
- custom-implementation
- deep-rl-class
model-index:
- name: pixelcopter_v3
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Pixelcopter-PLE-v0
type: Pixelcopter-PLE-v0
metrics:
- type: mean_reward
value: 57.70 +/- 36.28
name: mean_reward
verified: false
---
# **Reinforce** Agent playing **Pixelcopter-PLE-v0**
This is a trained model of a **Reinforce** agent playing **Pixelcopter-PLE-v0** .
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
|
JKuang96/pixelcopter | JKuang96 | 2024-02-01T06:47:51Z | 0 | 0 | null | [
"Pixelcopter-PLE-v0",
"reinforce",
"reinforcement-learning",
"custom-implementation",
"deep-rl-class",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-01-31T12:04:43Z | ---
tags:
- Pixelcopter-PLE-v0
- reinforce
- reinforcement-learning
- custom-implementation
- deep-rl-class
model-index:
- name: pixelcopter
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Pixelcopter-PLE-v0
type: Pixelcopter-PLE-v0
metrics:
- type: mean_reward
value: 38.30 +/- 54.66
name: mean_reward
verified: false
---
# **Reinforce** Agent playing **Pixelcopter-PLE-v0**
This is a trained model of a **Reinforce** agent playing **Pixelcopter-PLE-v0** .
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
|
worldboss/code-llama-7b-text-to-sql | worldboss | 2024-02-01T06:43:09Z | 0 | 0 | peft | [
"peft",
"tensorboard",
"safetensors",
"trl",
"sft",
"generated_from_trainer",
"dataset:generator",
"base_model:codellama/CodeLlama-7b-hf",
"base_model:adapter:codellama/CodeLlama-7b-hf",
"license:llama2",
"region:us"
]
| null | 2024-02-01T06:02:38Z | ---
license: llama2
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- generator
base_model: codellama/CodeLlama-7b-hf
model-index:
- name: code-llama-7b-text-to-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. -->
# code-llama-7b-text-to-sql
This model is a fine-tuned version of [codellama/CodeLlama-7b-hf](https://huggingface.co/codellama/CodeLlama-7b-hf) on the generator 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: 3
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 6
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 3
### Training results
### Framework versions
- PEFT 0.7.2.dev0
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1 |
bachbouch/w-3-qlora | bachbouch | 2024-02-01T06:39:33Z | 0 | 0 | peft | [
"peft",
"safetensors",
"mistral",
"arxiv:1910.09700",
"base_model:mistralai/Mistral-7B-Instruct-v0.2",
"base_model:adapter:mistralai/Mistral-7B-Instruct-v0.2",
"region:us"
]
| null | 2024-02-01T05:29:59Z | ---
library_name: peft
base_model: mistralai/Mistral-7B-Instruct-v0.2
---
# 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 0.8.2.dev0 |
mkdir700/v2-starcoderbase1b-personal-copilot-A100-40GB-colab | mkdir700 | 2024-02-01T06:38:24Z | 95 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"gpt_bigcode",
"text-generation",
"generated_from_trainer",
"base_model:bigcode/starcoderbase-1b",
"base_model:finetune:bigcode/starcoderbase-1b",
"license:bigcode-openrail-m",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-27T01:37:56Z | ---
license: bigcode-openrail-m
tags:
- generated_from_trainer
base_model: bigcode/starcoderbase-1b
model-index:
- name: v2-starcoderbase1b-personal-copilot-A100-40GB-colab
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. -->
# v2-starcoderbase1b-personal-copilot-A100-40GB-colab
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: nan
## 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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 30
- training_steps: 1000
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.0 | 0.1 | 100 | nan |
| 0.0 | 0.2 | 200 | nan |
| 0.0 | 0.3 | 300 | nan |
| 0.0 | 0.4 | 400 | nan |
| 0.0 | 0.5 | 500 | nan |
| 0.0 | 0.6 | 600 | nan |
| 0.0 | 0.7 | 700 | nan |
| 0.0 | 0.8 | 800 | nan |
| 0.0 | 0.9 | 900 | nan |
| 0.0 | 1.0 | 1000 | nan |
### Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
sleepyboi152/SolarMistral-10.7B | sleepyboi152 | 2024-02-01T06:37:35Z | 0 | 0 | null | [
"merge",
"mergekit",
"lazymergekit",
"cognitivecomputations/dolphin-2.6-mistral-7b",
"NousResearch/Nous-Hermes-2-SOLAR-10.7B",
"base_model:NousResearch/Nous-Hermes-2-SOLAR-10.7B",
"base_model:merge:NousResearch/Nous-Hermes-2-SOLAR-10.7B",
"base_model:cognitivecomputations/dolphin-2.6-mistral-7b",
"base_model:merge:cognitivecomputations/dolphin-2.6-mistral-7b",
"region:us"
]
| null | 2024-02-01T06:22:53Z | ---
tags:
- merge
- mergekit
- lazymergekit
- cognitivecomputations/dolphin-2.6-mistral-7b
- NousResearch/Nous-Hermes-2-SOLAR-10.7B
base_model:
- cognitivecomputations/dolphin-2.6-mistral-7b
- NousResearch/Nous-Hermes-2-SOLAR-10.7B
---
# SolarMistral-10.7B
SolarMistral-10.7B is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [cognitivecomputations/dolphin-2.6-mistral-7b](https://huggingface.co/cognitivecomputations/dolphin-2.6-mistral-7b)
* [NousResearch/Nous-Hermes-2-SOLAR-10.7B](https://huggingface.co/NousResearch/Nous-Hermes-2-SOLAR-10.7B)
## ๐งฉ Configuration
```yaml
slices:
- sources:
- model: cognitivecomputations/dolphin-2.6-mistral-7b
layer_range: [0, 32]
- model: NousResearch/Nous-Hermes-2-SOLAR-10.7B
layer_range: [0, 32]
merge_method: slerp
base_model: NousResearch/Nous-Hermes-2-SOLAR-10.7B
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
```
## ๐ป Usage
```python
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "sleepyboi152/SolarMistral-10.7B"
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"])
``` |
OpenDILabCommunity/TicTacToe-play-with-bot-SampledAlphaZero | OpenDILabCommunity | 2024-02-01T06:30:04Z | 0 | 0 | pytorch | [
"pytorch",
"deep-reinforcement-learning",
"reinforcement-learning",
"DI-engine",
"TicTacToe-play-with-bot",
"en",
"arxiv:2310.08348",
"license:apache-2.0",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T06:29:40Z | ---
language: en
license: apache-2.0
library_name: pytorch
tags:
- deep-reinforcement-learning
- reinforcement-learning
- DI-engine
- TicTacToe-play-with-bot
benchmark_name: OpenAI/Gym/Atari
task_name: TicTacToe-play-with-bot
pipeline_tag: reinforcement-learning
model-index:
- name: SampledAlphaZero
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: TicTacToe-play-with-bot
type: TicTacToe-play-with-bot
metrics:
- type: mean_reward
value: 0.3 +/- 0.64
name: mean_reward
---
# Play **TicTacToe-play-with-bot** with **SampledAlphaZero** Policy
## Model Description
<!-- Provide a longer summary of what this model is. -->
This implementation applies **SampledAlphaZero** to the OpenAI/Gym/Atari **TicTacToe-play-with-bot** environment using [LightZero](https://github.com/opendilab/LightZero) and [DI-engine](https://github.com/opendilab/di-engine).
**LightZero** is an efficient, easy-to-understand open-source toolkit that merges Monte Carlo Tree Search (MCTS) with Deep Reinforcement Learning (RL), simplifying their integration for developers and researchers. More details are in paper [LightZero: A Unified Benchmark for Monte Carlo Tree Search in General Sequential Decision Scenarios](https://huggingface.co/papers/2310.08348).
## Model Usage
### Install the Dependencies
<details close>
<summary>(Click for Details)</summary>
```shell
# install huggingface_ding
git clone https://github.com/opendilab/huggingface_ding.git
pip3 install -e ./huggingface_ding/
# install environment dependencies if needed
pip3 install DI-engine[common_env,video]
pip3 install LightZero
```
</details>
### Git Clone from Huggingface and Run the Model
<details close>
<summary>(Click for Details)</summary>
```shell
# running with trained model
python3 -u run.py
```
**run.py**
```python
from lzero.agent import SampledAlphaZeroAgent
from ding.config import Config
from easydict import EasyDict
import torch
# Pull model from files which are git cloned from huggingface
policy_state_dict = torch.load("pytorch_model.bin", map_location=torch.device("cpu"))
cfg = EasyDict(Config.file_to_dict("policy_config.py").cfg_dict)
# Instantiate the agent
agent = SampledAlphaZeroAgent(
env_id="TicTacToe-play-with-bot", exp_name="TicTacToe-play-with-bot-SampledAlphaZero", cfg=cfg.exp_config, policy_state_dict=policy_state_dict
)
# Continue training
agent.train(step=5000)
# Render the new agent performance
agent.deploy(enable_save_replay=True)
```
</details>
### Run Model by Using Huggingface_ding
<details close>
<summary>(Click for Details)</summary>
```shell
# running with trained model
python3 -u run.py
```
**run.py**
```python
from lzero.agent import SampledAlphaZeroAgent
from huggingface_ding import pull_model_from_hub
# Pull model from Hugggingface hub
policy_state_dict, cfg = pull_model_from_hub(repo_id="OpenDILabCommunity/TicTacToe-play-with-bot-SampledAlphaZero")
# Instantiate the agent
agent = SampledAlphaZeroAgent(
env_id="TicTacToe-play-with-bot", exp_name="TicTacToe-play-with-bot-SampledAlphaZero", cfg=cfg.exp_config, policy_state_dict=policy_state_dict
)
# Continue training
agent.train(step=5000)
# Render the new agent performance
agent.deploy(enable_save_replay=True)
```
</details>
## Model Training
### Train the Model and Push to Huggingface_hub
<details close>
<summary>(Click for Details)</summary>
```shell
#Training Your Own Agent
python3 -u train.py
```
**train.py**
```python
from lzero.agent import SampledAlphaZeroAgent
from huggingface_ding import push_model_to_hub
# Instantiate the agent
agent = SampledAlphaZeroAgent(env_id="TicTacToe-play-with-bot", exp_name="TicTacToe-play-with-bot-SampledAlphaZero")
# Train the agent
return_ = agent.train(step=int(500000))
# Push model to huggingface hub
push_model_to_hub(
agent=agent.best,
env_name="OpenAI/Gym/Atari",
task_name="TicTacToe-play-with-bot",
algo_name="SampledAlphaZero",
github_repo_url="https://github.com/opendilab/LightZero",
github_doc_model_url=None,
github_doc_env_url=None,
installation_guide='''
pip3 install DI-engine[common_env,video]
pip3 install LightZero
''',
usage_file_by_git_clone="./sampled_alphazero/tictactoe_play_with_bot_sampled_alphazero_deploy.py",
usage_file_by_huggingface_ding="./sampled_alphazero/tictactoe_play_with_bot_sampled_alphazero_download.py",
train_file="./sampled_alphazero/tictactoe_play_with_bot_sampled_alphazero.py",
repo_id="OpenDILabCommunity/TicTacToe-play-with-bot-SampledAlphaZero",
platform_info="[LightZero](https://github.com/opendilab/LightZero) and [DI-engine](https://github.com/opendilab/di-engine)",
model_description="**LightZero** is an efficient, easy-to-understand open-source toolkit that merges Monte Carlo Tree Search (MCTS) with Deep Reinforcement Learning (RL), simplifying their integration for developers and researchers. More details are in paper [LightZero: A Unified Benchmark for Monte Carlo Tree Search in General Sequential Decision Scenarios](https://huggingface.co/papers/2310.08348).",
create_repo=True
)
```
</details>
**Configuration**
<details close>
<summary>(Click for Details)</summary>
```python
exp_config = {
'main_config': {
'exp_name': 'TicTacToe-play-with-bot-SampledAlphaZero',
'seed': 0,
'env': {
'env_id': 'TicTacToe-play-with-bot',
'board_size': 3,
'battle_mode': 'play_with_bot_mode',
'bot_action_type': 'v0',
'channel_last': False,
'collector_env_num': 8,
'evaluator_env_num': 5,
'n_evaluator_episode': 5,
'manager': {
'shared_memory': False
},
'agent_vs_human': False,
'prob_random_agent': 0,
'prob_expert_agent': 0,
'scale': True,
'alphazero_mcts_ctree': False,
'save_replay_gif': False,
'replay_path_gif': './replay_gif'
},
'policy': {
'on_policy': False,
'cuda': True,
'multi_gpu': False,
'bp_update_sync': True,
'traj_len_inf': False,
'model': {
'observation_shape': [3, 3, 3],
'action_space_size': 9,
'num_res_blocks': 1,
'num_channels': 16,
'fc_value_layers': [8],
'fc_policy_layers': [8]
},
'torch_compile': False,
'tensor_float_32': False,
'sampled_algo': False,
'gumbel_algo': False,
'update_per_collect': 50,
'model_update_ratio': 0.1,
'batch_size': 256,
'optim_type': 'Adam',
'learning_rate': 0.003,
'weight_decay': 0.0001,
'momentum': 0.9,
'grad_clip_value': 0.5,
'value_weight': 1.0,
'collector_env_num': 8,
'evaluator_env_num': 5,
'lr_piecewise_constant_decay': False,
'threshold_training_steps_for_final_lr': 500000,
'manual_temperature_decay': False,
'threshold_training_steps_for_final_temperature': 100000,
'fixed_temperature_value': 0.25,
'mcts': {
'num_simulations': 25
},
'other': {
'replay_buffer': {
'replay_buffer_size': 1000000,
'save_episode': False
}
},
'cfg_type': 'AlphaZeroPolicyDict',
'mcts_ctree': False,
'simulation_env_name': 'tictactoe',
'simulation_env_config_type': 'play_with_bot',
'board_size': 3,
'entropy_weight': 0.0,
'n_episode': 8,
'eval_freq': 2000
},
'wandb_logger': {
'gradient_logger': False,
'video_logger': False,
'plot_logger': False,
'action_logger': False,
'return_logger': False
}
},
'create_config': {
'env': {
'type': 'tictactoe',
'import_names': ['zoo.board_games.tictactoe.envs.tictactoe_env']
},
'env_manager': {
'type': 'subprocess'
},
'policy': {
'type': 'alphazero',
'import_names': ['lzero.policy.alphazero']
},
'collector': {
'type': 'episode_alphazero',
'import_names': ['lzero.worker.alphazero_collector']
},
'evaluator': {
'type': 'alphazero',
'import_names': ['lzero.worker.alphazero_evaluator']
}
}
}
```
</details>
**Training Procedure**
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
- **Weights & Biases (wandb):** [monitor link](<TODO>)
## Model Information
<!-- Provide the basic links for the model. -->
- **Github Repository:** [repo link](https://github.com/opendilab/LightZero)
- **Doc**: [Algorithm link](<TODO>)
- **Configuration:** [config link](https://huggingface.co/OpenDILabCommunity/TicTacToe-play-with-bot-SampledAlphaZero/blob/main/policy_config.py)
- **Demo:** [video](https://huggingface.co/OpenDILabCommunity/TicTacToe-play-with-bot-SampledAlphaZero/blob/main/replay.mp4)
<!-- Provide the size information for the model. -->
- **Parameters total size:** 51.13 KB
- **Last Update Date:** 2024-02-01
## Environments
<!-- Address questions around what environment the model is intended to be trained and deployed at, including the necessary information needed to be provided for future users. -->
- **Benchmark:** OpenAI/Gym/Atari
- **Task:** TicTacToe-play-with-bot
- **Gym version:** 0.25.1
- **DI-engine version:** v0.5.0
- **PyTorch version:** 2.0.1+cu117
- **Doc**: [Environments link](<TODO>)
|
cchoi1022/librispeech-100h-supervised_not_mine | cchoi1022 | 2024-02-01T06:28:03Z | 118 | 0 | transformers | [
"transformers",
"safetensors",
"wav2vec2",
"automatic-speech-recognition",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| automatic-speech-recognition | 2024-02-01T06:22:48Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
cchoi1022/my-test-model | cchoi1022 | 2024-02-01T06:21:19Z | 119 | 0 | transformers | [
"transformers",
"safetensors",
"wav2vec2",
"automatic-speech-recognition",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| automatic-speech-recognition | 2024-02-01T06:19:48Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
Hui-1/speaker_recognition-campplus-zh-en-16k-base | Hui-1 | 2024-02-01T06:20:57Z | 0 | 0 | null | [
"CAM++",
"3D-Speaker",
"Speaker Recognition & Verification",
"Chinese & English",
"zh",
"en",
"dataset:3D-Speaker",
"dataset:voxceleb",
"dataset:cnceleb",
"license:apache-2.0",
"region:us"
]
| null | 2024-01-31T09:42:22Z | ---
language:
- zh
- en
tags:
- CAM++
- 3D-Speaker
- Speaker Recognition & Verification
- Chinese & English
license: apache-2.0
datasets:
- 3D-Speaker
- voxceleb
- cnceleb
metrics:
- EER
- minDCF
---
# CAM++: A Fast and Efficient Speaker Recognition Network
CAM++ is a fast and efficient network based on a densely-connected time delay neural network (D-TDNN). This repository provides tools for extracting speaker embeddings and performing speaker verification tasks with a pretrained CAM++ model. It has been trained on a large-scale training dataset, which includes Chinese and English corpora.
|
LanguageBind/MoE-LLaVA-Phi2-Pretrain | LanguageBind | 2024-02-01T06:10:06Z | 14 | 0 | transformers | [
"transformers",
"llava_phi",
"text-generation",
"custom_code",
"arxiv:2401.15947",
"arxiv:2311.10122",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-31T11:32:06Z | ---
license: apache-2.0
---
<p align="center">
<img src="https://s11.ax1x.com/2023/12/28/piqvDMV.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/abs/2401.15947">MoE-LLaVA: Mixture of Experts for Large Vision-Language Models</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
<h5 align="center">
</h5>
## ๐ฐ News
* **[2024.01.30]** The [paper](https://arxiv.org/abs/2401.15947) is released.
* **[2024.01.27]** ๐ค[Hugging Face demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) and **all codes & datasets** are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
MoE-LLaVA shows excellent performance in multi-modal learning.
### ๐ฅ High performance, but with fewer parameters
- with just **3B sparsely activated parameters**, MoE-LLaVA demonstrates performance comparable to the LLaVA-1.5-7B on various visual understanding datasets and even surpasses the LLaVA-1.5-13B in object hallucination benchmarks.
### ๐ Simple baseline, learning multi-modal interactions with sparse pathways.
- With the addition of **a simple MoE tuning stage**, we can complete the training of MoE-LLaVA on **8 V100 GPUs** within 2 days.
## ๐ค Demo
### Gradio Web UI
Highly recommend trying out our web demo by the following command, which incorporates all features currently supported by MoE-LLaVA. We also provide [online demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) in Huggingface Spaces.
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e"
# use qwen
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e"
# use stablelm
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e"
```
### CLI Inference
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e" --image-file "image.jpg"
# use qwen
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e" --image-file "image.jpg"
# use stablelm
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e" --image-file "image.jpg"
```
## ๐ณ Model Zoo
| Model | LLM | Checkpoint | Avg | VQAv2 | GQA | VizWiz | SQA | T-VQA | POPE | MM-Bench| LLaVA-Bench-Wild | MM-Vet |
|----------|-----------|-----------|---|---|---|---|---|---|---|---|---|---|
| MoE-LLaVA-1.6Bร4-Top2 | 1.6B | [LanguageBind/MoE-LLaVA-StableLM-1.6B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-StableLM-1.6B-4e) | 60.0 | 76.0 | 60.4 | 37.2 | 62.6 | 47.8 | 84.3 | 59.4 | 85.9 | 26.1 |
| MoE-LLaVA-1.8Bร4-Top2 | 1.8B | [LanguageBind/MoE-LLaVA-Qwen-1.8B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Qwen-1.8B-4e) | 60.2 | 76.2 | 61.5 | 32.6 | 63.1 | 48.0 | 87.0 | 59.6 | 88.7 | 25.3 |
| MoE-LLaVA-2.7Bร4-Top2 | 2.7B | [LanguageBind/MoE-LLaVA-Phi2-2.7B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Phi2-2.7B-4e) | 63.9 | 77.1 | 61.1 | 43.4 | 68.7 | 50.2 | 85.0 | 65.5 | 93.2 | 31.1 |
<!--
| LLaVA-1.5 | 7B | [liuhaotian/llava-v1.5-7b](https://huggingface.co/liuhaotian/llava-v1.5-7b) | 62.0 | 78.5 | 62.0 | 50.0 | 66.8 | 58.2 | 85.9 | 64.3 | 31.1 |
| LLaVA-1.5 | 13B | [liuhaotian/llava-v1.5-13b](https://huggingface.co/liuhaotian/llava-v1.5-13b) | 64.9 | 80.0 | 63.3 | 53.6 | 71.6 | 61.3 | 85.9 | 67.7 | 36.1 |
-->
## โ๏ธ Requirements and Installation
* Python >= 3.10
* Pytorch == 2.0.1
* CUDA Version >= 11.7
* **Transformers == 4.36.2**
* **Tokenizers==0.15.1**
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/MoE-LLaVA
cd MoE-LLaVA
conda create -n moellava python=3.10 -y
conda activate moellava
pip install --upgrade pip # enable PEP 660 support
pip install -e .
pip install -e ".[train]"
pip install flash-attn --no-build-isolation
# Below are optional. For Qwen model.
git clone https://github.com/Dao-AILab/flash-attention
cd flash-attention && pip install .
# Below are optional. Installing them might be slow.
# pip install csrc/layer_norm
# If the version of flash-attn is higher than 2.1.1, the following is not needed.
# pip install csrc/rotary
```
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN.md](docs/TRAIN.md) & [EVAL.md](docs/EVAL.md).
## ๐ก Customizing your MoE-LLaVA
The instruction is in [CUSTOM.md](docs/CUSTOM.md).
## ๐ Visualization
The instruction is in [VISUALIZATION.md](docs/VISUALIZATION.md).
## ๐ค API
**We open source all codes.** If you want to load the model (e.g. ```LanguageBind/MoE-LLaVA```) on local, you can use the following code snippets.
**Using the following command to run the code.**
```bash
deepspeed predict.py
```
```python
import torch
from moellava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
from moellava.conversation import conv_templates, SeparatorStyle
from moellava.model.builder import load_pretrained_model
from moellava.utils import disable_torch_init
from moellava.mm_utils import tokenizer_image_token, get_model_name_from_path, KeywordsStoppingCriteria
def main():
disable_torch_init()
image = 'moellava/serve/examples/extreme_ironing.jpg'
inp = 'What is unusual about this image?'
model_path = 'LanguageBind/MoE-LLaVA-Phi2-2.7B-4e' # LanguageBind/MoE-LLaVA-Qwen-1.8B-4e or LanguageBind/MoE-LLaVA-StableLM-1.6B-4e
device = 'cuda'
load_4bit, load_8bit = False, False # FIXME: Deepspeed support 4bit or 8bit?
model_name = get_model_name_from_path(model_path)
tokenizer, model, processor, context_len = load_pretrained_model(model_path, None, model_name, load_8bit, load_4bit, device=device)
image_processor = processor['image']
conv_mode = "phi" # qwen or stablelm
conv = conv_templates[conv_mode].copy()
roles = conv.roles
image_tensor = image_processor.preprocess(image, return_tensors='pt')['pixel_values'].to(model.device, dtype=torch.float16)
print(f"{roles[1]}: {inp}")
inp = DEFAULT_IMAGE_TOKEN + '\n' + inp
conv.append_message(conv.roles[0], inp)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).cuda()
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=image_tensor,
do_sample=True,
temperature=0.2,
max_new_tokens=1024,
use_cache=True,
stopping_criteria=[stopping_criteria])
outputs = tokenizer.decode(output_ids[0, input_ids.shape[1]:], skip_special_tokens=True).strip()
print(outputs)
if __name__ == '__main__':
main()
```
## ๐ Related Projects
* [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) This framework empowers the model to efficiently utilize the united visual tokens.
* [LanguageBind](https://github.com/PKU-YuanGroup/LanguageBind) An open source five modalities language-based retrieval framework.
## ๐ Acknowledgement
* [LLaVA](https://github.com/haotian-liu/LLaVA) The codebase we built upon and it is an efficient large language and vision assistant.
## ๐ License
* The majority of this project is released under the Apache 2.0 license as found in the [LICENSE](https://github.com/PKU-YuanGroup/MoE-LLaVA/blob/main/LICENSE) file.
* The service is a research preview intended for non-commercial use only, subject to the model [License](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md) of LLaMA, [Terms of Use](https://openai.com/policies/terms-of-use) of the data generated by OpenAI, and [Privacy Practices](https://chrome.google.com/webstore/detail/sharegpt-share-your-chatg/daiacboceoaocpibfodeljbdfacokfjb) of ShareGPT. Please contact us if you find any potential violation.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{lin2024moellava,
title={MoE-LLaVA: Mixture of Experts for Large Vision-Language Models},
author={Bin Lin and Zhenyu Tang and Yang Ye and Jiaxi Cui and Bin Zhu and Peng Jin and Junwu Zhang and Munan Ning and Li Yuan},
year={2024},
eprint={2401.15947},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
```BibTeX
@article{lin2023video,
title={Video-LLaVA: Learning United Visual Representation by Alignment Before Projection},
author={Lin, Bin and Zhu, Bin and Ye, Yang and Ning, Munan and Jin, Peng and Yuan, Li},
journal={arXiv preprint arXiv:2311.10122},
year={2023}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/MoE-LLaVA&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/MoE-LLaVA/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/MoE-LLaVA" />
</a>
|
LanguageBind/MoE-LLaVA-Phi2-384-Pretrain | LanguageBind | 2024-02-01T06:09:39Z | 14 | 0 | transformers | [
"transformers",
"llava_phi",
"text-generation",
"custom_code",
"arxiv:2401.15947",
"arxiv:2311.10122",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-31T11:34:09Z | ---
license: apache-2.0
---
<p align="center">
<img src="https://s11.ax1x.com/2023/12/28/piqvDMV.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/abs/2401.15947">MoE-LLaVA: Mixture of Experts for Large Vision-Language Models</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
<h5 align="center">
</h5>
## ๐ฐ News
* **[2024.01.30]** The [paper](https://arxiv.org/abs/2401.15947) is released.
* **[2024.01.27]** ๐ค[Hugging Face demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) and **all codes & datasets** are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
MoE-LLaVA shows excellent performance in multi-modal learning.
### ๐ฅ High performance, but with fewer parameters
- with just **3B sparsely activated parameters**, MoE-LLaVA demonstrates performance comparable to the LLaVA-1.5-7B on various visual understanding datasets and even surpasses the LLaVA-1.5-13B in object hallucination benchmarks.
### ๐ Simple baseline, learning multi-modal interactions with sparse pathways.
- With the addition of **a simple MoE tuning stage**, we can complete the training of MoE-LLaVA on **8 V100 GPUs** within 2 days.
## ๐ค Demo
### Gradio Web UI
Highly recommend trying out our web demo by the following command, which incorporates all features currently supported by MoE-LLaVA. We also provide [online demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) in Huggingface Spaces.
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e"
# use qwen
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e"
# use stablelm
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e"
```
### CLI Inference
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e" --image-file "image.jpg"
# use qwen
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e" --image-file "image.jpg"
# use stablelm
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e" --image-file "image.jpg"
```
## ๐ณ Model Zoo
| Model | LLM | Checkpoint | Avg | VQAv2 | GQA | VizWiz | SQA | T-VQA | POPE | MM-Bench| LLaVA-Bench-Wild | MM-Vet |
|----------|-----------|-----------|---|---|---|---|---|---|---|---|---|---|
| MoE-LLaVA-1.6Bร4-Top2 | 1.6B | [LanguageBind/MoE-LLaVA-StableLM-1.6B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-StableLM-1.6B-4e) | 60.0 | 76.0 | 60.4 | 37.2 | 62.6 | 47.8 | 84.3 | 59.4 | 85.9 | 26.1 |
| MoE-LLaVA-1.8Bร4-Top2 | 1.8B | [LanguageBind/MoE-LLaVA-Qwen-1.8B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Qwen-1.8B-4e) | 60.2 | 76.2 | 61.5 | 32.6 | 63.1 | 48.0 | 87.0 | 59.6 | 88.7 | 25.3 |
| MoE-LLaVA-2.7Bร4-Top2 | 2.7B | [LanguageBind/MoE-LLaVA-Phi2-2.7B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Phi2-2.7B-4e) | 63.9 | 77.1 | 61.1 | 43.4 | 68.7 | 50.2 | 85.0 | 65.5 | 93.2 | 31.1 |
<!--
| LLaVA-1.5 | 7B | [liuhaotian/llava-v1.5-7b](https://huggingface.co/liuhaotian/llava-v1.5-7b) | 62.0 | 78.5 | 62.0 | 50.0 | 66.8 | 58.2 | 85.9 | 64.3 | 31.1 |
| LLaVA-1.5 | 13B | [liuhaotian/llava-v1.5-13b](https://huggingface.co/liuhaotian/llava-v1.5-13b) | 64.9 | 80.0 | 63.3 | 53.6 | 71.6 | 61.3 | 85.9 | 67.7 | 36.1 |
-->
## โ๏ธ Requirements and Installation
* Python >= 3.10
* Pytorch == 2.0.1
* CUDA Version >= 11.7
* **Transformers == 4.36.2**
* **Tokenizers==0.15.1**
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/MoE-LLaVA
cd MoE-LLaVA
conda create -n moellava python=3.10 -y
conda activate moellava
pip install --upgrade pip # enable PEP 660 support
pip install -e .
pip install -e ".[train]"
pip install flash-attn --no-build-isolation
# Below are optional. For Qwen model.
git clone https://github.com/Dao-AILab/flash-attention
cd flash-attention && pip install .
# Below are optional. Installing them might be slow.
# pip install csrc/layer_norm
# If the version of flash-attn is higher than 2.1.1, the following is not needed.
# pip install csrc/rotary
```
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN.md](docs/TRAIN.md) & [EVAL.md](docs/EVAL.md).
## ๐ก Customizing your MoE-LLaVA
The instruction is in [CUSTOM.md](docs/CUSTOM.md).
## ๐ Visualization
The instruction is in [VISUALIZATION.md](docs/VISUALIZATION.md).
## ๐ค API
**We open source all codes.** If you want to load the model (e.g. ```LanguageBind/MoE-LLaVA```) on local, you can use the following code snippets.
**Using the following command to run the code.**
```bash
deepspeed predict.py
```
```python
import torch
from moellava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
from moellava.conversation import conv_templates, SeparatorStyle
from moellava.model.builder import load_pretrained_model
from moellava.utils import disable_torch_init
from moellava.mm_utils import tokenizer_image_token, get_model_name_from_path, KeywordsStoppingCriteria
def main():
disable_torch_init()
image = 'moellava/serve/examples/extreme_ironing.jpg'
inp = 'What is unusual about this image?'
model_path = 'LanguageBind/MoE-LLaVA-Phi2-2.7B-4e' # LanguageBind/MoE-LLaVA-Qwen-1.8B-4e or LanguageBind/MoE-LLaVA-StableLM-1.6B-4e
device = 'cuda'
load_4bit, load_8bit = False, False # FIXME: Deepspeed support 4bit or 8bit?
model_name = get_model_name_from_path(model_path)
tokenizer, model, processor, context_len = load_pretrained_model(model_path, None, model_name, load_8bit, load_4bit, device=device)
image_processor = processor['image']
conv_mode = "phi" # qwen or stablelm
conv = conv_templates[conv_mode].copy()
roles = conv.roles
image_tensor = image_processor.preprocess(image, return_tensors='pt')['pixel_values'].to(model.device, dtype=torch.float16)
print(f"{roles[1]}: {inp}")
inp = DEFAULT_IMAGE_TOKEN + '\n' + inp
conv.append_message(conv.roles[0], inp)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).cuda()
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=image_tensor,
do_sample=True,
temperature=0.2,
max_new_tokens=1024,
use_cache=True,
stopping_criteria=[stopping_criteria])
outputs = tokenizer.decode(output_ids[0, input_ids.shape[1]:], skip_special_tokens=True).strip()
print(outputs)
if __name__ == '__main__':
main()
```
## ๐ Related Projects
* [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) This framework empowers the model to efficiently utilize the united visual tokens.
* [LanguageBind](https://github.com/PKU-YuanGroup/LanguageBind) An open source five modalities language-based retrieval framework.
## ๐ Acknowledgement
* [LLaVA](https://github.com/haotian-liu/LLaVA) The codebase we built upon and it is an efficient large language and vision assistant.
## ๐ License
* The majority of this project is released under the Apache 2.0 license as found in the [LICENSE](https://github.com/PKU-YuanGroup/MoE-LLaVA/blob/main/LICENSE) file.
* The service is a research preview intended for non-commercial use only, subject to the model [License](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md) of LLaMA, [Terms of Use](https://openai.com/policies/terms-of-use) of the data generated by OpenAI, and [Privacy Practices](https://chrome.google.com/webstore/detail/sharegpt-share-your-chatg/daiacboceoaocpibfodeljbdfacokfjb) of ShareGPT. Please contact us if you find any potential violation.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{lin2024moellava,
title={MoE-LLaVA: Mixture of Experts for Large Vision-Language Models},
author={Bin Lin and Zhenyu Tang and Yang Ye and Jiaxi Cui and Bin Zhu and Peng Jin and Junwu Zhang and Munan Ning and Li Yuan},
year={2024},
eprint={2401.15947},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
```BibTeX
@article{lin2023video,
title={Video-LLaVA: Learning United Visual Representation by Alignment Before Projection},
author={Lin, Bin and Zhu, Bin and Ye, Yang and Ning, Munan and Jin, Peng and Yuan, Li},
journal={arXiv preprint arXiv:2311.10122},
year={2023}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/MoE-LLaVA&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/MoE-LLaVA/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/MoE-LLaVA" />
</a>
|
LanguageBind/MoE-LLaVA-Qwen-1.8B-4e | LanguageBind | 2024-02-01T06:09:13Z | 199 | 13 | transformers | [
"transformers",
"pytorch",
"moe_llava_qwen",
"text-generation",
"custom_code",
"arxiv:2401.15947",
"arxiv:2311.10122",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-23T13:50:43Z | ---
license: apache-2.0
---
<p align="center">
<img src="https://s11.ax1x.com/2023/12/28/piqvDMV.png" width="250" style="margin-bottom: 0.2;"/>
<p>
<h2 align="center"> <a href="https://arxiv.org/abs/2401.15947">MoE-LLaVA: Mixture of Experts for Large Vision-Language Models</a></h2>
<h5 align="center"> If you like our project, please give us a star โญ on GitHub for latest update. </h2>
<h5 align="center">
</h5>
## ๐ฐ News
* **[2024.01.30]** The [paper](https://arxiv.org/abs/2401.15947) is released.
* **[2024.01.27]** ๐ค[Hugging Face demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) and **all codes & datasets** are available now! Welcome to **watch** ๐ this repository for the latest updates.
## ๐ฎ Highlights
MoE-LLaVA shows excellent performance in multi-modal learning.
### ๐ฅ High performance, but with fewer parameters
- with just **3B sparsely activated parameters**, MoE-LLaVA demonstrates performance comparable to the LLaVA-1.5-7B on various visual understanding datasets and even surpasses the LLaVA-1.5-13B in object hallucination benchmarks.
### ๐ Simple baseline, learning multi-modal interactions with sparse pathways.
- With the addition of **a simple MoE tuning stage**, we can complete the training of MoE-LLaVA on **8 V100 GPUs** within 2 days.
## ๐ค Demo
### Gradio Web UI
Highly recommend trying out our web demo by the following command, which incorporates all features currently supported by MoE-LLaVA. We also provide [online demo](https://huggingface.co/spaces/LanguageBind/MoE-LLaVA) in Huggingface Spaces.
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e"
# use qwen
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e"
# use stablelm
deepspeed --include localhost:0 moellava/serve/gradio_web_server.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e"
```
### CLI Inference
```bash
# use phi2
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Phi2-2.7B-4e" --image-file "image.jpg"
# use qwen
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-Qwen-1.8B-4e" --image-file "image.jpg"
# use stablelm
deepspeed --include localhost:0 moellava/serve/cli.py --model-path "LanguageBind/MoE-LLaVA-StableLM-1.6B-4e" --image-file "image.jpg"
```
## ๐ณ Model Zoo
| Model | LLM | Checkpoint | Avg | VQAv2 | GQA | VizWiz | SQA | T-VQA | POPE | MM-Bench| LLaVA-Bench-Wild | MM-Vet |
|----------|-----------|-----------|---|---|---|---|---|---|---|---|---|---|
| MoE-LLaVA-1.6Bร4-Top2 | 1.6B | [LanguageBind/MoE-LLaVA-StableLM-1.6B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-StableLM-1.6B-4e) | 60.0 | 76.0 | 60.4 | 37.2 | 62.6 | 47.8 | 84.3 | 59.4 | 85.9 | 26.1 |
| MoE-LLaVA-1.8Bร4-Top2 | 1.8B | [LanguageBind/MoE-LLaVA-Qwen-1.8B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Qwen-1.8B-4e) | 60.2 | 76.2 | 61.5 | 32.6 | 63.1 | 48.0 | 87.0 | 59.6 | 88.7 | 25.3 |
| MoE-LLaVA-2.7Bร4-Top2 | 2.7B | [LanguageBind/MoE-LLaVA-Phi2-2.7B-4e](https://huggingface.co/LanguageBind/MoE-LLaVA-Phi2-2.7B-4e) | 63.9 | 77.1 | 61.1 | 43.4 | 68.7 | 50.2 | 85.0 | 65.5 | 93.2 | 31.1 |
<!--
| LLaVA-1.5 | 7B | [liuhaotian/llava-v1.5-7b](https://huggingface.co/liuhaotian/llava-v1.5-7b) | 62.0 | 78.5 | 62.0 | 50.0 | 66.8 | 58.2 | 85.9 | 64.3 | 31.1 |
| LLaVA-1.5 | 13B | [liuhaotian/llava-v1.5-13b](https://huggingface.co/liuhaotian/llava-v1.5-13b) | 64.9 | 80.0 | 63.3 | 53.6 | 71.6 | 61.3 | 85.9 | 67.7 | 36.1 |
-->
## โ๏ธ Requirements and Installation
* Python >= 3.10
* Pytorch == 2.0.1
* CUDA Version >= 11.7
* **Transformers == 4.36.2**
* **Tokenizers==0.15.1**
* Install required packages:
```bash
git clone https://github.com/PKU-YuanGroup/MoE-LLaVA
cd MoE-LLaVA
conda create -n moellava python=3.10 -y
conda activate moellava
pip install --upgrade pip # enable PEP 660 support
pip install -e .
pip install -e ".[train]"
pip install flash-attn --no-build-isolation
# Below are optional. For Qwen model.
git clone https://github.com/Dao-AILab/flash-attention
cd flash-attention && pip install .
# Below are optional. Installing them might be slow.
# pip install csrc/layer_norm
# If the version of flash-attn is higher than 2.1.1, the following is not needed.
# pip install csrc/rotary
```
## ๐๏ธ Training & Validating
The training & validating instruction is in [TRAIN.md](docs/TRAIN.md) & [EVAL.md](docs/EVAL.md).
## ๐ก Customizing your MoE-LLaVA
The instruction is in [CUSTOM.md](docs/CUSTOM.md).
## ๐ Visualization
The instruction is in [VISUALIZATION.md](docs/VISUALIZATION.md).
## ๐ค API
**We open source all codes.** If you want to load the model (e.g. ```LanguageBind/MoE-LLaVA```) on local, you can use the following code snippets.
**Using the following command to run the code.**
```bash
deepspeed predict.py
```
```python
import torch
from moellava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
from moellava.conversation import conv_templates, SeparatorStyle
from moellava.model.builder import load_pretrained_model
from moellava.utils import disable_torch_init
from moellava.mm_utils import tokenizer_image_token, get_model_name_from_path, KeywordsStoppingCriteria
def main():
disable_torch_init()
image = 'moellava/serve/examples/extreme_ironing.jpg'
inp = 'What is unusual about this image?'
model_path = 'LanguageBind/MoE-LLaVA-Phi2-2.7B-4e' # LanguageBind/MoE-LLaVA-Qwen-1.8B-4e or LanguageBind/MoE-LLaVA-StableLM-1.6B-4e
device = 'cuda'
load_4bit, load_8bit = False, False # FIXME: Deepspeed support 4bit or 8bit?
model_name = get_model_name_from_path(model_path)
tokenizer, model, processor, context_len = load_pretrained_model(model_path, None, model_name, load_8bit, load_4bit, device=device)
image_processor = processor['image']
conv_mode = "phi" # qwen or stablelm
conv = conv_templates[conv_mode].copy()
roles = conv.roles
image_tensor = image_processor.preprocess(image, return_tensors='pt')['pixel_values'].to(model.device, dtype=torch.float16)
print(f"{roles[1]}: {inp}")
inp = DEFAULT_IMAGE_TOKEN + '\n' + inp
conv.append_message(conv.roles[0], inp)
conv.append_message(conv.roles[1], None)
prompt = conv.get_prompt()
input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).cuda()
stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
keywords = [stop_str]
stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)
with torch.inference_mode():
output_ids = model.generate(
input_ids,
images=image_tensor,
do_sample=True,
temperature=0.2,
max_new_tokens=1024,
use_cache=True,
stopping_criteria=[stopping_criteria])
outputs = tokenizer.decode(output_ids[0, input_ids.shape[1]:], skip_special_tokens=True).strip()
print(outputs)
if __name__ == '__main__':
main()
```
## ๐ Related Projects
* [Video-LLaVA](https://github.com/PKU-YuanGroup/Video-LLaVA) This framework empowers the model to efficiently utilize the united visual tokens.
* [LanguageBind](https://github.com/PKU-YuanGroup/LanguageBind) An open source five modalities language-based retrieval framework.
## ๐ Acknowledgement
* [LLaVA](https://github.com/haotian-liu/LLaVA) The codebase we built upon and it is an efficient large language and vision assistant.
## ๐ License
* The majority of this project is released under the Apache 2.0 license as found in the [LICENSE](https://github.com/PKU-YuanGroup/MoE-LLaVA/blob/main/LICENSE) file.
* The service is a research preview intended for non-commercial use only, subject to the model [License](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md) of LLaMA, [Terms of Use](https://openai.com/policies/terms-of-use) of the data generated by OpenAI, and [Privacy Practices](https://chrome.google.com/webstore/detail/sharegpt-share-your-chatg/daiacboceoaocpibfodeljbdfacokfjb) of ShareGPT. Please contact us if you find any potential violation.
## โ๏ธ Citation
If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:.
```BibTeX
@misc{lin2024moellava,
title={MoE-LLaVA: Mixture of Experts for Large Vision-Language Models},
author={Bin Lin and Zhenyu Tang and Yang Ye and Jiaxi Cui and Bin Zhu and Peng Jin and Junwu Zhang and Munan Ning and Li Yuan},
year={2024},
eprint={2401.15947},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
```BibTeX
@article{lin2023video,
title={Video-LLaVA: Learning United Visual Representation by Alignment Before Projection},
author={Lin, Bin and Zhu, Bin and Ye, Yang and Ning, Munan and Jin, Peng and Yuan, Li},
journal={arXiv preprint arXiv:2311.10122},
year={2023}
}
```
## โจ Star History
[](https://star-history.com/#PKU-YuanGroup/MoE-LLaVA&Date)
## ๐ค Contributors
<a href="https://github.com/PKU-YuanGroup/MoE-LLaVA/graphs/contributors">
<img src="https://contrib.rocks/image?repo=PKU-YuanGroup/MoE-LLaVA" />
</a>
|
k-seungri/k_whisper_tokenizer | k-seungri | 2024-02-01T05:58:50Z | 0 | 0 | transformers | [
"transformers",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| null | 2024-01-30T09:15:25Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
arpanl/fine-tuned | arpanl | 2024-02-01T05:52:31Z | 176 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"vit",
"image-classification",
"generated_from_trainer",
"dataset:imagefolder",
"base_model:google/vit-base-patch16-224",
"base_model:finetune:google/vit-base-patch16-224",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| image-classification | 2024-02-01T05:51:36Z | ---
license: apache-2.0
base_model: google/vit-base-patch16-224
tags:
- generated_from_trainer
datasets:
- imagefolder
model-index:
- name: fine-tuned
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. -->
# fine-tuned
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the imagefolder 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: 8
- 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 4.37.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
bartowski/DistilabelBeagle14-7B-exl2 | bartowski | 2024-02-01T05:48:08Z | 6 | 0 | null | [
"merge",
"mergekit",
"lazymergekit",
"dpo",
"rlhf",
"rlaif",
"distilabel",
"text-generation",
"base_model:mlabonne/Beagle14-7B",
"base_model:finetune:mlabonne/Beagle14-7B",
"license:cc-by-nc-4.0",
"region:us"
]
| text-generation | 2024-02-01T05:32:02Z | ---
license: cc-by-nc-4.0
base_model: mlabonne/Beagle14-7B
tags:
- merge
- mergekit
- lazymergekit
- dpo
- rlhf
- rlaif
- distilabel
quantized_by: bartowski
pipeline_tag: text-generation
---
## Exllama v2 Quantizations of DistilabelBeagle14-7B
Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.12">turboderp's ExLlamaV2 v0.0.12</a> for quantization.
# The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)
Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
Original model: https://huggingface.co/argilla/DistilabelBeagle14-7B
| Branch | Bits | lm_head bits | VRAM (4k) | VRAM (16k) | VRAM (32k) | Description |
| ----- | ---- | ------- | ------ | ------ | ------ | ------------ |
| [8_0](https://huggingface.co/Bartowski/DistilabelBeagle14-7B-exl2/tree/8_0) | 8.0 | 8.0 | 8.4 GB | 9.8 GB | 11.8 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
| [6_5](https://huggingface.co/Bartowski/DistilabelBeagle14-7B-exl2/tree/6_5) | 6.5 | 8.0 | 7.2 GB | 8.6 GB | 10.6 GB | Very similar to 8.0, good tradeoff of size vs performance, **recommended**. |
| [5_0](https://huggingface.co/Bartowski/DistilabelBeagle14-7B-exl2/tree/5_0) | 5.0 | 6.0 | 6.0 GB | 7.4 GB | 9.4 GB | Slightly lower quality vs 6.5, but usable on 8GB cards. |
| [4_25](https://huggingface.co/Bartowski/DistilabelBeagle14-7B-exl2/tree/4_25) | 4.25 | 6.0 | 5.3 GB | 6.7 GB | 8.7 GB | GPTQ equivalent bits per weight, slightly higher quality. |
| [3_5](https://huggingface.co/Bartowski/DistilabelBeagle14-7B-exl2/tree/3_5) | 3.5 | 6.0 | 4.7 GB | 6.1 GB | 8.1 GB | Lower quality, only use if you have to. |
## Download instructions
With git:
```shell
git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/DistilabelBeagle14-7B-exl2 DistilabelBeagle14-7B-exl2-6_5
```
With huggingface hub (credit to TheBloke for instructions):
```shell
pip3 install huggingface-hub
```
To download the `main` (only useful if you only care about measurement.json) branch to a folder called `DistilabelBeagle14-7B-exl2`:
```shell
mkdir DistilabelBeagle14-7B-exl2
huggingface-cli download bartowski/DistilabelBeagle14-7B-exl2 --local-dir DistilabelBeagle14-7B-exl2 --local-dir-use-symlinks False
```
To download from a different branch, add the `--revision` parameter:
Linux:
```shell
mkdir DistilabelBeagle14-7B-exl2-6_5
huggingface-cli download bartowski/DistilabelBeagle14-7B-exl2 --revision 6_5 --local-dir DistilabelBeagle14-7B-exl2-6_5 --local-dir-use-symlinks False
```
Windows (which apparently doesn't like _ in folders sometimes?):
```shell
mkdir DistilabelBeagle14-7B-exl2-6.5
huggingface-cli download bartowski/DistilabelBeagle14-7B-exl2 --revision 6_5 --local-dir DistilabelBeagle14-7B-exl2-6.5 --local-dir-use-symlinks False
```
Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski |
Tsuka25/w2v-bert-2.0-mongolian-colab-CV16.0 | Tsuka25 | 2024-02-01T05:46:28Z | 60 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"wav2vec2-bert",
"automatic-speech-recognition",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| automatic-speech-recognition | 2024-01-30T08:01:07Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
Joshua-Abok/flan-t5-large-dialogsum_samsum_v2 | Joshua-Abok | 2024-02-01T05:37:00Z | 92 | 0 | transformers | [
"transformers",
"safetensors",
"t5",
"text2text-generation",
"arxiv:1910.09700",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text2text-generation | 2024-02-01T05:30:42Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
LoneStriker/DistilabelBeagle14-7B-GGUF | LoneStriker | 2024-02-01T05:31:25Z | 7 | 0 | null | [
"gguf",
"merge",
"mergekit",
"lazymergekit",
"dpo",
"rlhf",
"rlaif",
"distilabel",
"arxiv:1910.09700",
"base_model:mlabonne/Beagle14-7B",
"base_model:quantized:mlabonne/Beagle14-7B",
"license:cc-by-nc-4.0",
"endpoints_compatible",
"region:us",
"conversational"
]
| null | 2024-02-01T05:08:56Z | ---
license: cc-by-nc-4.0
base_model: mlabonne/Beagle14-7B
tags:
- merge
- mergekit
- lazymergekit
- dpo
- rlhf
- rlaif
- distilabel
---
# Model Card for Model ID
This is a preference tuned version of `mlabonne/Beagle14-7B` using a mix of Argilla's orca pairs and a new upcoming multi-turn dpo dataset.
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** Argilla
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** English
- **License:** cc-by-nc-4.0
- **Finetuned from model [optional]:** mlabonne/Beagle14-7B
### 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]
|
cloudyu/Phoenix_DPO_60B | cloudyu | 2024-02-01T05:28:31Z | 111 | 1 | transformers | [
"transformers",
"safetensors",
"mixtral",
"text-generation",
"yi",
"moe",
"license:other",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-26T08:55:55Z | ---
license: other
tags:
- yi
- moe
license_name: yi-license
license_link: https://huggingface.co/01-ai/Yi-34B-200K/blob/main/LICENSE
---
this is a DPO fine-tuned MoE model with 60B parameter.
```
DPO Trainer
TRL supports the DPO Trainer for training language models from preference data, as described in the paper Direct Preference Optimization: Your Language Model is Secretly a Reward Model by Rafailov et al., 2023.
```
GGUF format is ready at [cloudyu/Phoenix_DPO_60B_gguf](https://huggingface.co/cloudyu/Phoenix_DPO_60B_gguf)
|
vivekdugale/en_pipeline | vivekdugale | 2024-02-01T05:18:14Z | 2 | 0 | spacy | [
"spacy",
"token-classification",
"en",
"model-index",
"region:us"
]
| token-classification | 2024-01-31T13:26:33Z | ---
tags:
- spacy
- token-classification
language:
- en
model-index:
- name: en_pipeline
results:
- task:
name: NER
type: token-classification
metrics:
- name: NER Precision
type: precision
value: 0.7554585153
- name: NER Recall
type: recall
value: 0.8277511962
- name: NER F Score
type: f_score
value: 0.7899543379
---
| Feature | Description |
| --- | --- |
| **Name** | `en_pipeline` |
| **Version** | `0.0.0` |
| **spaCy** | `>=3.7.2,<3.8.0` |
| **Default Pipeline** | `transformer`, `ner` |
| **Components** | `transformer`, `ner` |
| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) |
| **Sources** | n/a |
| **License** | n/a |
| **Author** | [n/a]() |
### Label Scheme
<details>
<summary>View label scheme (1 labels for 1 components)</summary>
| Component | Labels |
| --- | --- |
| **`ner`** | `SECTOR` |
</details>
### Accuracy
| Type | Score |
| --- | --- |
| `ENTS_F` | 79.00 |
| `ENTS_P` | 75.55 |
| `ENTS_R` | 82.78 |
| `TRANSFORMER_LOSS` | 10044.45 |
| `NER_LOSS` | 47764.82 | |
mindlywork/Maskot1 | mindlywork | 2024-02-01T05:13:47Z | 2 | 0 | diffusers | [
"diffusers",
"text-to-image",
"stable-diffusion",
"lora",
"template:sd-lora",
"base_model:stabilityai/stable-diffusion-xl-base-1.0",
"base_model:adapter:stabilityai/stable-diffusion-xl-base-1.0",
"license:cc",
"region:us"
]
| text-to-image | 2024-02-01T05:11:06Z | ---
tags:
- text-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
widget:
- text: Maskot1
output:
url: images/out-0 (12).png
base_model: stabilityai/stable-diffusion-xl-base-1.0
instance_prompt: Maskot1
license: cc
---
# Maskot1
<Gallery />
## Model description
Maskot1
## Trigger words
You should use `Maskot1` to trigger the image generation.
## Download model
Weights for this model are available in Safetensors format.
[Download](/dasdsff/Maskot1/tree/main) them in the Files & versions tab.
|
KuriT/dqn-SpaceInvadersNoFrameskip-v4 | KuriT | 2024-02-01T05:03:47Z | 0 | 0 | stable-baselines3 | [
"stable-baselines3",
"SpaceInvadersNoFrameskip-v4",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T05:03:11Z | ---
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: 646.50 +/- 168.11
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 KuriT -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 KuriT -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 KuriT
```
## 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'}
```
|
octnn/ppo-SnowballTarget | octnn | 2024-02-01T05:00:27Z | 4 | 0 | ml-agents | [
"ml-agents",
"tensorboard",
"onnx",
"SnowballTarget",
"deep-reinforcement-learning",
"reinforcement-learning",
"ML-Agents-SnowballTarget",
"region:us"
]
| reinforcement-learning | 2024-02-01T05:00:24Z | ---
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 ๐ถ 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: octnn/ppo-SnowballTarget
3. Step 2: Select your *.nn /*.onnx file
4. Click on Watch the agent play ๐
|
XiaQiang/dqn-SpaceInvadersNoFrameskip-v4 | XiaQiang | 2024-02-01T04:58:16Z | 0 | 0 | stable-baselines3 | [
"stable-baselines3",
"SpaceInvadersNoFrameskip-v4",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T04:57:35Z | ---
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: 610.50 +/- 224.01
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 XiaQiang -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 XiaQiang -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 XiaQiang
```
## 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'}
```
|
dhanikitkat/topic-modelling-indo | dhanikitkat | 2024-02-01T04:54:18Z | 119 | 0 | transformers | [
"transformers",
"pytorch",
"bert",
"text-classification",
"topic modelling indo",
"id",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-classification | 2024-02-01T04:45:08Z | ---
license: mit
language:
- id
pipeline_tag: text-classification
tags:
- topic modelling indo
--- |
jlbaker361/ddpo-stability-dcgan | jlbaker361 | 2024-02-01T04:51:43Z | 0 | 1 | diffusers | [
"diffusers",
"safetensors",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
]
| text-to-image | 2024-02-01T04:47:56Z | ---
{}
---
# DDPO trained model
num_epochs=10
train_gradient_accumulation_steps=1
sample_num_steps=30
sample_batch_size=16
train_batch_size=16
sample_num_batches_per_epoch=32
based off of stabilityai/stable-diffusion-2-base
and then trained off of None
|
jan-hq/stealth-rag-v1 | jan-hq | 2024-02-01T04:34:45Z | 3 | 0 | peft | [
"peft",
"safetensors",
"mistral",
"alignment-handbook",
"generated_from_trainer",
"trl",
"sft",
"dataset:jan-hq/rag_dataset_1200_binarized",
"dataset:jan-hq/rag_dataset_12000_binarized",
"dataset:jan-hq/rag_hallucination_dataset_1000_binarized",
"dataset:jan-hq/rag_full_20000_binarized",
"dataset:jan-hq/bagel_sft_binarized",
"dataset:jan-hq/dolphin_binarized",
"dataset:jan-hq/openhermes_binarized",
"base_model:jan-hq/stealth-v1.3",
"base_model:adapter:jan-hq/stealth-v1.3",
"license:apache-2.0",
"4-bit",
"bitsandbytes",
"region:us"
]
| null | 2024-02-01T02:21:12Z | ---
license: apache-2.0
library_name: peft
tags:
- alignment-handbook
- generated_from_trainer
- trl
- sft
- generated_from_trainer
datasets:
- jan-hq/rag_dataset_1200_binarized
- jan-hq/rag_dataset_12000_binarized
- jan-hq/rag_hallucination_dataset_1000_binarized
- jan-hq/rag_full_20000_binarized
- jan-hq/bagel_sft_binarized
- jan-hq/dolphin_binarized
- jan-hq/openhermes_binarized
base_model: jan-hq/stealth-v1.3
model-index:
- name: stealth-rag-v1
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. -->
# stealth-rag-v1
This model is a fine-tuned version of [jan-hq/stealth-v1.3](https://huggingface.co/jan-hq/stealth-v1.3) on the jan-hq/rag_dataset_1200_binarized, the jan-hq/rag_dataset_12000_binarized, the jan-hq/rag_hallucination_dataset_1000_binarized, the jan-hq/rag_full_20000_binarized, the jan-hq/bagel_sft_binarized, the jan-hq/dolphin_binarized and the jan-hq/openhermes_binarized datasets.
It achieves the following results on the evaluation set:
- Loss: 1.3883
## 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: 2
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- total_eval_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
### Training results
### Framework versions
- PEFT 0.7.1
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.0 |
andysalerno/rainbowfish-v2 | andysalerno | 2024-02-01T04:22:03Z | 6 | 0 | transformers | [
"transformers",
"safetensors",
"mistral",
"text-generation",
"conversational",
"arxiv:1910.09700",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T04:09:32Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
cloudyu/Mixtral_7Bx4_MOE_DPO | cloudyu | 2024-02-01T04:19:02Z | 15 | 1 | transformers | [
"transformers",
"safetensors",
"mixtral",
"text-generation",
"license:cc-by-nc-4.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-01-18T06:40:57Z | ---
license: cc-by-nc-4.0
---
* [This is DPO improved version of cloudyu/Mixtral_7Bx4_MOE_24B](https://huggingface.co/cloudyu/Mixtral_7Bx4_MOE_24B)
* [DPO Trainer](https://huggingface.co/docs/trl/main/en/dpo_trainer)
* Metrics improved by DPO


|
devjwsong/q-FrozenLake-v1-4x4-noSlippery | devjwsong | 2024-02-01T04:12:58Z | 0 | 0 | null | [
"FrozenLake-v1-4x4-no_slippery",
"q-learning",
"reinforcement-learning",
"custom-implementation",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T04:12:56Z | ---
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="devjwsong/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"])
```
|
zhaoxinwind/QA_model | zhaoxinwind | 2024-02-01T03:52:04Z | 111 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"distilbert",
"question-answering",
"generated_from_trainer",
"base_model:distilbert/distilbert-base-uncased",
"base_model:finetune:distilbert/distilbert-base-uncased",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
]
| question-answering | 2024-01-29T11:54:53Z | ---
license: apache-2.0
base_model: distilbert-base-uncased
tags:
- generated_from_trainer
model-index:
- name: QA_model
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. -->
# QA_model
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0017
## 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: 16
- eval_batch_size: 16
- 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 |
|:-------------:|:-----:|:----:|:---------------:|
| 0.0855 | 1.0 | 1202 | 0.0812 |
| 0.0312 | 2.0 | 2404 | 0.0017 |
| 0.0005 | 3.0 | 3606 | 0.0017 |
### Framework versions
- Transformers 4.37.2
- Pytorch 2.1.2+cpu
- Datasets 2.16.1
- Tokenizers 0.15.1
|
octnn/Reinforce-Pixelcopter-PLE-v0 | octnn | 2024-02-01T03:51:01Z | 0 | 0 | null | [
"Pixelcopter-PLE-v0",
"reinforce",
"reinforcement-learning",
"custom-implementation",
"deep-rl-class",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-01-31T03:08:08Z | ---
tags:
- Pixelcopter-PLE-v0
- reinforce
- reinforcement-learning
- custom-implementation
- deep-rl-class
model-index:
- name: Reinforce-Pixelcopter-PLE-v0
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Pixelcopter-PLE-v0
type: Pixelcopter-PLE-v0
metrics:
- type: mean_reward
value: 24.00 +/- 20.29
name: mean_reward
verified: false
---
# **Reinforce** Agent playing **Pixelcopter-PLE-v0**
This is a trained model of a **Reinforce** agent playing **Pixelcopter-PLE-v0** .
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
|
alpindale/miquella-120b-gguf | alpindale | 2024-02-01T03:50:53Z | 49 | 17 | null | [
"gguf",
"merge",
"endpoints_compatible",
"region:us"
]
| null | 2024-02-01T02:00:08Z | ---
tags:
- merge
---
# Miquella 120B GGUF
GGUF quantized weights for [miquella-120b](https://huggingface.co/alpindale/miquella-120b). Contains *all* quants.
I used Importance Matrices generated from Q8_0 quant of the model. The dataset used for that was random junk
for optimal quality.
Due to the limitations of HF's file size, the larger files were split into multiple chunks. Instructions below.
## Linux
Example uses Q3_K_L. Replace the names appropriately for your quant of choice.
```sh
cat miquella-120b.Q3_K_L.gguf_part_* > miquella-120b.Q3_K_L.gguf && rm miquella-120b.Q3_K_L.gguf_part_*
```
## Windows
Example uses Q3_K_L. Replace the names appropriately for your quant of choice.
```sh
COPY /B miquella-120b.Q3_K_L.gguf_part_aa + miquella-120b.Q3_K_L.gguf_part_ab miquella-120b.Q3_K_L.gguf
```
Then delete the two splits. |
whythoname/TheWondermixV11 | whythoname | 2024-02-01T03:39:12Z | 1 | 0 | diffusers | [
"diffusers",
"region:us"
]
| null | 2024-02-01T03:15:59Z | ---
library_name: diffusers
--- |
tr-aravindan/output_emotion | tr-aravindan | 2024-02-01T03:28:35Z | 2 | 0 | peft | [
"peft",
"safetensors",
"trl",
"sft",
"generated_from_trainer",
"dataset:emotion",
"base_model:openai-community/gpt2",
"base_model:adapter:openai-community/gpt2",
"license:mit",
"region:us"
]
| null | 2024-01-23T08:20:28Z | ---
license: mit
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- emotion
base_model: gpt2
model-index:
- name: output_emotion
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. -->
# output_emotion
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 7.0852
## 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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 8.5945 | 2.0 | 1800 | 7.5670 |
| 7.7948 | 4.0 | 3600 | 7.0852 |
### Framework versions
- PEFT 0.7.1
- Transformers 4.36.2
- Pytorch 2.0.0
- Datasets 2.15.0
- Tokenizers 0.15.0 |
tr-aravindan/output | tr-aravindan | 2024-02-01T03:28:35Z | 3 | 0 | peft | [
"peft",
"safetensors",
"generated_from_trainer",
"dataset:emotion",
"base_model:openai-community/gpt2",
"base_model:adapter:openai-community/gpt2",
"license:mit",
"region:us"
]
| null | 2024-01-20T21:31:04Z | ---
license: mit
library_name: peft
tags:
- generated_from_trainer
datasets:
- emotion
base_model: gpt2
model-index:
- name: output
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. -->
# output
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 3.8303
## 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: 0.001
- 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: linear
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| No log | 0.25 | 250 | 4.8185 |
| No log | 0.5 | 500 | 4.3814 |
| No log | 0.75 | 750 | 4.1230 |
| No log | 1.0 | 1000 | 4.0088 |
| No log | 1.25 | 1250 | 3.9536 |
| No log | 1.5 | 1500 | 3.9208 |
| No log | 1.75 | 1750 | 3.8946 |
| 4.644 | 2.0 | 2000 | 3.8799 |
| 4.644 | 2.25 | 2250 | 3.8651 |
| 4.644 | 2.5 | 2500 | 3.8552 |
| 4.644 | 2.75 | 2750 | 3.8464 |
| 4.644 | 3.0 | 3000 | 3.8399 |
| 4.644 | 3.25 | 3250 | 3.8364 |
| 4.644 | 3.5 | 3500 | 3.8333 |
| 4.644 | 3.75 | 3750 | 3.8311 |
| 4.0742 | 4.0 | 4000 | 3.8303 |
### Framework versions
- PEFT 0.7.1
- Transformers 4.36.2
- Pytorch 2.1.2
- Datasets 2.15.0
- Tokenizers 0.15.1 |
AIFT/AIFT-instruct-dpo-v1.3-42dot_LLM-SFT-1.3B | AIFT | 2024-02-01T03:25:12Z | 278 | 0 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"license:cc-by-sa-4.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T01:24:11Z | ---
license: cc-by-sa-4.0
---
<h1>AIFT-instruct-42dot_LLM-SFT-1.3B</h1>
<b><ํ์ต ๋ฐ์ดํฐ ๊ตฌ์ถ></b>
<br>
kyujinpy ๋์ด ๊ณต๊ฐํ์ KOR-OpenOrca-Platypus ๋ฐ์ดํฐ๋ฅผ ์ผ๋ถ ์ญ์ (์ํ๋ง) ๋ฐ ์ ์ ์์
์งํํ์ฌ ํ์ฉ.
๊ทธ ์ดํ ํด๋น ๋ฐ์ดํฐ๋ค์ ๋ณด๋ฉฐ ๊ด๋ จ ํ์คํฌ๋ฅผ ์ถ์ถํ์๊ณ ์ด๋ฅผ ๊ธฐ๋ฐ์ผ๋ก
ํด๋น ํ์คํฌ์ ๋ง์ถฐ์ NLP ๊ด๋ จ ์คํ์์ค ๋ฐ์ดํฐ๋ฅผ ํ์ฉํ์ฌ ํ์ต๋ฐ์ดํฐ๋ฅผ ์์ฒด์ ์ผ๋ก
์ญ์ฌ, ๊ณผํ, ์ํ, ๊ธฐ๊ณ๋
ํด, ๋ฆฌ๋ทฐ ๋ถ์ ๋ฌธ์ ๋ฅผ gpt๋ฅผ ํตํด์ ๊ตฌ์ถํ์๊ณ ,
aihub ์ผ๋ฐ์์ ๋ฐ ๊ธฐ๊ณ๋
ํด ๋ฐ์ดํฐ๋ฅผ ํ์ฉํ์ฌ ์ถ๊ฐ๋ก ํ์ต ๋ฐ์ดํฐ๋ฅผ ๊ตฌ์ถ(ํํ์ ๊ด๋ จ, ๊ธฐ๊ณ๋
ํด ๊ด๋ จ ๋ฐ ์์ฝ)
๊ฐ์ข
๋ธ๋ก๊ทธ์์ ์ญ์ฌ ๋ฐ ์์ ํด์ฆ๋ฅผ ์ฌ๋์ด ์ง์ ํ์ต๋ฐ์ดํฐ ํํ๋ก ๋ณ๊ฒฝ
AI2AI Challenge ๋ฐ์ดํฐ ํํ๋ฅผ ๋ณด๊ณ gpt๋ฅผ ํตํด ์ด๋ฑ ์์ค์ ๊ณผํ ์ํ ๋ฌธ์ ์ ํ์ ์ ์ 500๋ฌธ์
์์ด ๋ฒ์ญ ๋ฐ์ดํฐ ์ํ/ํ์ ๋ฐ์ดํฐ ํ์ต ๋ฐ์ดํฐ๋ก ํ์ฉ ์งํ
์ด ๋ฐ์ดํฐ 4๋ง๊ฐ ์ ๋ ์ฌ์ฉํ์์ต๋๋ค.
<br>
<br>
+ TruthfulQA ๊ด๋ จ ๋ฌธ์ ์ถ๊ฐ๋ฅผ ์งํํ์์ต๋๋ค.(์์ค ๊ด๋ จ ์ฐธ๊ฑฐ์ง ๋ฌธ์ )
+ ๊ธฐ๊ณ๋
ํด ๊ด๋ จ ํ์ต ๋ฐ์ดํฐ๋ฅผ ChatGPT๋ฅผ ํตํด์ ๋ต๋ณ์ ์ป์ด ํ์ต
+ ๋ฌธ๋ฒ๊ด๋ จ ํ์ต ๋ฐ์ดํฐ
<br>
<br>
DPO ๋ฐ์ดํฐ์
<br>
ko-HH-RLHF ๋ฐ์ดํฐ์ chosen ๋ฐ์ดํฐ๋ฅผ gpt-3.5-turbo๋ฅผ ํตํด ์ฌ ์์ฑํ์ฌ ํ์ต์ ํ์ฉํ์์ต๋๋ค.
๋ํ TruthfulQA ๋ฐ์ดํฐ ์ฝ 1200๊ฑด์ ์์ฒด ์ ์ํ์์ต๋๋ค.
<br>
###ํ์ต ๋ฐ์ดํฐ ํ์ผ์ ๋น๊ณต๊ฐ์
๋๋ค.
<br>
<๋ชจ๋ธ>
<br>
42dot์์ ๊ณต๊ฐํ 42dot_LLM-SFT-1.3B์ ๋ฒ ์ด์ค ๋ชจ๋ธ๋ก ํ์ฌ ํ์ต ์งํํ์์ต๋๋ค.
<br>
<br>
<br>
<b><ํ์ต></b>
<br>
ํ์ต์ LoRA๋ฅผ ์ฌ์ฉํ์ฌ A100 40G *2์์ ํ์ต์ ์งํํ์์ต๋๋ค.
|
erik1126/Enet | erik1126 | 2024-02-01T03:24:11Z | 0 | 0 | null | [
"license:apache-2.0",
"region:us"
]
| null | 2024-01-23T08:16:36Z | ---
license: apache-2.0
---
# Model for signature verification
Don't use it in production. This is just for an experiment. |
DataGenius/MyLinkedlnProPic | DataGenius | 2024-02-01T03:19:57Z | 1 | 1 | 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 | 2024-02-01T03:19:01Z |
---
base_model: stabilityai/stable-diffusion-xl-base-1.0
instance_prompt: Hazz photos
tags:
- text-to-image
- diffusers
- autotrain
inference: true
---
# DreamBooth trained by AutoTrain
Text encoder was not trained.
|
YaHi/teacher_electra_small_building_binary | YaHi | 2024-02-01T03:09:56Z | 89 | 0 | transformers | [
"transformers",
"pytorch",
"electra",
"text-classification",
"generated_from_trainer",
"base_model:google/electra-small-discriminator",
"base_model:finetune:google/electra-small-discriminator",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| text-classification | 2024-01-19T02:08:36Z | ---
license: apache-2.0
base_model: google/electra-small-discriminator
tags:
- generated_from_trainer
model-index:
- name: teacher_electra_small_building_binary
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. -->
# teacher_electra_small_building_binary
This model is a fine-tuned version of [google/electra-small-discriminator](https://huggingface.co/google/electra-small-discriminator) 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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 1022
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 4
### Training results
### Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0
- Datasets 2.12.0
- Tokenizers 0.13.2
|
chathuranga-jayanath/codet5-small-v11 | chathuranga-jayanath | 2024-02-01T03:00:57Z | 6 | 0 | transformers | [
"transformers",
"tensorboard",
"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 | 2024-01-31T18:09:43Z | ---
license: apache-2.0
base_model: Salesforce/codet5-small
tags:
- generated_from_trainer
model-index:
- name: codet5-small-v11
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. -->
# codet5-small-v11
This model is a fine-tuned version of [Salesforce/codet5-small](https://huggingface.co/Salesforce/codet5-small) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1410
- Bleu Score: 0.0004
- Gen Len: 13.1271
## Model description
Trained:
- on: chathuranga-jayanath/formatted-selfapr-train-data
- prompt: selfapr train data format -> [BUG]...[CONTEXT]...[CLASS]...[ERROR] <fe> <ce>
## 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: 30
- eval_batch_size: 30
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu Score | Gen Len |
|:-------------:|:-----:|:-----:|:---------------:|:----------:|:-------:|
| 0.2246 | 1.0 | 10656 | 0.1778 | 0.0004 | 13.1185 |
| 0.1915 | 2.0 | 21312 | 0.1489 | 0.0004 | 13.1311 |
| 0.1623 | 3.0 | 31968 | 0.1410 | 0.0004 | 13.1271 |
### Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
Anguuuuus/enhanced_german | Anguuuuus | 2024-02-01T02:54:49Z | 146 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"wav2vec2",
"audio-classification",
"generated_from_trainer",
"base_model:facebook/wav2vec2-base",
"base_model:finetune:facebook/wav2vec2-base",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
]
| audio-classification | 2024-01-29T05:57:45Z | ---
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: enhanced_german
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. -->
# enhanced_german
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6891
- Accuracy: 0.6667
## 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: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log | 1.0 | 1 | 0.7047 | 0.3333 |
| No log | 2.0 | 2 | 0.6962 | 0.5 |
| No log | 3.0 | 3 | 0.6878 | 0.5833 |
| No log | 4.0 | 4 | 0.6962 | 0.6667 |
| No log | 5.0 | 5 | 0.6954 | 0.6667 |
| No log | 6.0 | 6 | 0.6951 | 0.5833 |
| No log | 7.0 | 7 | 0.6922 | 0.5833 |
| No log | 8.0 | 8 | 0.6907 | 0.5833 |
| No log | 9.0 | 9 | 0.6899 | 0.5833 |
| 0.3093 | 10.0 | 10 | 0.6891 | 0.6667 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
cloudyu/Yi-34Bx2-MoE-60B-GGUF | cloudyu | 2024-02-01T02:49:08Z | 16 | 4 | null | [
"gguf",
"endpoints_compatible",
"region:us",
"conversational"
]
| null | 2024-02-01T02:22:00Z | ---
tags:
- gguf
---
## Description
This repo contains GGUF format model files for [cloudyu/Yi-34Bx2-MoE-60B](https://huggingface.co/cloudyu/Yi-34Bx2-MoE-60B).
<!-- 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.
### How to run GGUF with llama.cpp on an A10 (24G vram)
```
git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp/
make LLAMA_CUBLAS=1
./main --model ./cloudyu_Yi-34Bx2-MoE-60B_Q3_K_XS.gguf -p "what is biggest animal?" -i -ngl 36
``` |
alk/phi2-dolly-sum-finetune | alk | 2024-02-01T02:46:01Z | 3 | 0 | peft | [
"peft",
"safetensors",
"generated_from_trainer",
"base_model:microsoft/phi-2",
"base_model:adapter:microsoft/phi-2",
"license:mit",
"region:us"
]
| null | 2024-02-01T02:45:56Z | ---
license: mit
library_name: peft
tags:
- generated_from_trainer
base_model: microsoft/phi-2
model-index:
- name: phi2-dolly-sum-finetune
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. -->
# phi2-dolly-sum-finetune
This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.0505
## 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: 2
- 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: 1
- training_steps: 500
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 2.9329 | 0.05 | 25 | 2.4178 |
| 2.4832 | 0.09 | 50 | 2.1541 |
| 2.1688 | 0.14 | 75 | 2.0774 |
| 2.2247 | 0.18 | 100 | 2.0725 |
| 2.225 | 0.23 | 125 | 2.0652 |
| 2.2217 | 0.27 | 150 | 2.0635 |
| 2.2282 | 0.32 | 175 | 2.0611 |
| 2.1104 | 0.37 | 200 | 2.0608 |
| 2.1583 | 0.41 | 225 | 2.0569 |
| 2.1197 | 0.46 | 250 | 2.0565 |
| 2.1257 | 0.5 | 275 | 2.0559 |
| 2.0018 | 0.55 | 300 | 2.0512 |
| 2.0203 | 0.6 | 325 | 2.0546 |
| 2.1332 | 0.64 | 350 | 2.0519 |
| 2.1585 | 0.69 | 375 | 2.0503 |
| 2.1287 | 0.73 | 400 | 2.0510 |
| 2.1431 | 0.78 | 425 | 2.0515 |
| 2.1601 | 0.82 | 450 | 2.0522 |
| 2.088 | 0.87 | 475 | 2.0481 |
| 2.0462 | 0.92 | 500 | 2.0505 |
### Framework versions
- PEFT 0.8.1
- Transformers 4.37.2
- Pytorch 2.2.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1 |
devjwsong/ppo-Huggy | devjwsong | 2024-02-01T02:38:36Z | 0 | 0 | ml-agents | [
"ml-agents",
"tensorboard",
"onnx",
"Huggy",
"deep-reinforcement-learning",
"reinforcement-learning",
"ML-Agents-Huggy",
"region:us"
]
| reinforcement-learning | 2024-02-01T02:38:31Z | ---
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 ๐ถ 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: devjwsong/ppo-Huggy
3. Step 2: Select your *.nn /*.onnx file
4. Click on Watch the agent play ๐
|
son-of-man/HoloViolet-7B-test4 | son-of-man | 2024-02-01T02:36:38Z | 6 | 0 | transformers | [
"transformers",
"safetensors",
"mistral",
"text-generation",
"merge",
"mergekit",
"lazymergekit",
"KoboldAI/Mistral-7B-Holodeck-1",
"FPHam/Sydney_Pirate_Mistral_7b",
"base_model:FPHam/Sydney_Pirate_Mistral_7b",
"base_model:merge:FPHam/Sydney_Pirate_Mistral_7b",
"base_model:KoboldAI/Mistral-7B-Holodeck-1",
"base_model:merge:KoboldAI/Mistral-7B-Holodeck-1",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T02:33:05Z | ---
tags:
- merge
- mergekit
- lazymergekit
- KoboldAI/Mistral-7B-Holodeck-1
- FPHam/Sydney_Pirate_Mistral_7b
base_model:
- KoboldAI/Mistral-7B-Holodeck-1
- FPHam/Sydney_Pirate_Mistral_7b
---
# HoloViolet-7B-test4
HoloViolet-7B-test4 is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [KoboldAI/Mistral-7B-Holodeck-1](https://huggingface.co/KoboldAI/Mistral-7B-Holodeck-1)
* [FPHam/Sydney_Pirate_Mistral_7b](https://huggingface.co/FPHam/Sydney_Pirate_Mistral_7b)
## ๐งฉ Configuration
```yaml
merge_method: dare_ties
base_model: GreenNode/GreenNode-mini-7B-multilingual-v1olet
models:
- model: GreenNode/GreenNode-mini-7B-multilingual-v1olet
- model: KoboldAI/Mistral-7B-Holodeck-1
parameters:
weight: 0.42
- model: FPHam/Sydney_Pirate_Mistral_7b
parameters:
weight: 0.21
dtype: bfloat16
```
## ๐ป Usage
```python
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "son-of-man/HoloViolet-7B-test4"
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"])
``` |
jlbaker361/dcgan-gpu-wikiart25 | jlbaker361 | 2024-02-01T02:33:59Z | 0 | 0 | null | [
"region:us"
]
| null | 2024-01-25T06:24:49Z | ---
{}
---
Creative Adversarial Network
epochs: 3
dataset jlbaker361/wikiart-balanced25
n classes 27
batch_size 4
images where resized to 768
and then center cropped to: 512
used clip=False
discriminator parameters:
init_dim: 32
final_dim 512
generator parameters:
input noise_dim: 100
|
omartariq612/quran-whisper-large-v3-epoch-2 | omartariq612 | 2024-02-01T02:25:07Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| null | 2024-02-01T02:25:03Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
rhplus0831/maid-yuzu-v2-alter | rhplus0831 | 2024-02-01T02:19:34Z | 5 | 0 | transformers | [
"transformers",
"safetensors",
"mixtral",
"text-generation",
"mergekit",
"merge",
"base_model:smelborp/MixtralOrochi8x7B",
"base_model:merge:smelborp/MixtralOrochi8x7B",
"base_model:ycros/BagelMIsteryTour-v2-8x7B",
"base_model:merge:ycros/BagelMIsteryTour-v2-8x7B",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T01:52:19Z | ---
base_model:
- smelborp/MixtralOrochi8x7B
- ycros/BagelMIsteryTour-v2-8x7B
tags:
- mergekit
- merge
---
# maid-yuzu-v2-alter
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
This model is an experiment to combine two models I liked.
## Merge Details
### Merge Method
This model was merged using the SLERP merge method.
### Models Merged
The following models were included in the merge:
* [smelborp/MixtralOrochi8x7B](https://huggingface.co/smelborp/MixtralOrochi8x7B)
* [ycros/BagelMIsteryTour-v2-8x7B](https://huggingface.co/ycros/BagelMIsteryTour-v2-8x7B)
### Configuration
The following YAML configuration was used to produce this model:
```yaml
base_model:
model:
path: smelborp/MixtralOrochi8x7B
dtype: bfloat16
merge_method: slerp
parameters:
t:
- value: 0.5
slices:
- sources:
- layer_range: [0, 32]
model:
model:
path: smelborp/MixtralOrochi8x7B
- layer_range: [0, 32]
model:
model:
path: ycros/BagelMIsteryTour-v2-8x7B
```
|
blzncz/segformer-finetuned-4ss1st3r_s3gs3m_24Jan_negro-10k-steps | blzncz | 2024-02-01T02:09:47Z | 178 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"segformer",
"image-segmentation",
"vision",
"generated_from_trainer",
"license:other",
"endpoints_compatible",
"region:us"
]
| image-segmentation | 2024-01-31T10:49:13Z | ---
license: other
tags:
- image-segmentation
- vision
- generated_from_trainer
model-index:
- name: segformer-finetuned-4ss1st3r_s3gs3m_24Jan_negro-10k-steps
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. -->
# segformer-finetuned-4ss1st3r_s3gs3m_24Jan_negro-10k-steps
This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the blzncz/4ss1st3r_s3gs3m_24Jan_negro dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2142
- Mean Iou: 0.5724
- Mean Accuracy: 0.7571
- Overall Accuracy: 0.9468
- Accuracy Bg: nan
- Accuracy Fallo cohesivo: 0.9826
- Accuracy Fallo malla: 0.7246
- Accuracy Fallo adhesivo: 0.9679
- Accuracy Fallo burbuja: 0.3533
- Iou Bg: 0.0
- Iou Fallo cohesivo: 0.9368
- Iou Fallo malla: 0.6678
- Iou Fallo adhesivo: 0.9310
- Iou Fallo burbuja: 0.3263
## 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: 6e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 1337
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: polynomial
- training_steps: 10000
### Training results
| Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Accuracy Bg | Accuracy Fallo cohesivo | Accuracy Fallo malla | Accuracy Fallo adhesivo | Accuracy Fallo burbuja | Iou Bg | Iou Fallo cohesivo | Iou Fallo malla | Iou Fallo adhesivo | Iou Fallo burbuja |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:-------------:|:----------------:|:-----------:|:-----------------------:|:--------------------:|:-----------------------:|:----------------------:|:------:|:------------------:|:---------------:|:------------------:|:-----------------:|
| 0.1902 | 1.0 | 219 | 0.2615 | 0.5072 | 0.7247 | 0.9038 | nan | 0.9203 | 0.7991 | 0.9401 | 0.2393 | 0.0 | 0.8853 | 0.5556 | 0.9006 | 0.1944 |
| 0.1367 | 2.0 | 438 | 0.2067 | 0.5492 | 0.7602 | 0.9293 | nan | 0.9599 | 0.7487 | 0.9317 | 0.4004 | 0.0 | 0.9160 | 0.6120 | 0.8988 | 0.3195 |
| 0.1066 | 3.0 | 657 | 0.1963 | 0.5659 | 0.7814 | 0.9313 | nan | 0.9520 | 0.8022 | 0.9545 | 0.4169 | 0.0 | 0.9175 | 0.6270 | 0.9267 | 0.3584 |
| 0.1102 | 4.0 | 876 | 0.1595 | 0.5782 | 0.7756 | 0.9444 | nan | 0.9727 | 0.7669 | 0.9693 | 0.3934 | 0.0 | 0.9336 | 0.6828 | 0.9326 | 0.3422 |
| 0.1114 | 5.0 | 1095 | 0.1678 | 0.5772 | 0.7950 | 0.9378 | nan | 0.9619 | 0.7778 | 0.9756 | 0.4648 | 0.0 | 0.9255 | 0.6534 | 0.9151 | 0.3922 |
| 0.0897 | 6.0 | 1314 | 0.1726 | 0.5811 | 0.7976 | 0.9420 | nan | 0.9701 | 0.7598 | 0.9723 | 0.4881 | 0.0 | 0.9307 | 0.6613 | 0.9170 | 0.3965 |
| 0.0788 | 7.0 | 1533 | 0.2096 | 0.5491 | 0.7253 | 0.9342 | nan | 0.9898 | 0.5936 | 0.9381 | 0.3797 | 0.0 | 0.9235 | 0.5698 | 0.9149 | 0.3374 |
| 0.0788 | 8.0 | 1752 | 0.1574 | 0.5774 | 0.7733 | 0.9465 | nan | 0.9726 | 0.7858 | 0.9675 | 0.3673 | 0.0 | 0.9359 | 0.6914 | 0.9264 | 0.3331 |
| 0.0855 | 9.0 | 1971 | 0.1970 | 0.5406 | 0.7141 | 0.9380 | nan | 0.9866 | 0.6305 | 0.9708 | 0.2687 | 0.0 | 0.9274 | 0.5984 | 0.9224 | 0.2548 |
| 0.0761 | 10.0 | 2190 | 0.1903 | 0.5564 | 0.7479 | 0.9382 | nan | 0.9746 | 0.7050 | 0.9737 | 0.3383 | 0.0 | 0.9268 | 0.6272 | 0.9182 | 0.3098 |
| 0.0686 | 11.0 | 2409 | 0.1910 | 0.5562 | 0.7435 | 0.9393 | nan | 0.9827 | 0.6605 | 0.9738 | 0.3572 | 0.0 | 0.9285 | 0.6156 | 0.9209 | 0.3160 |
| 0.062 | 12.0 | 2628 | 0.2038 | 0.5453 | 0.7399 | 0.9334 | nan | 0.9728 | 0.6739 | 0.9811 | 0.3317 | 0.0 | 0.9214 | 0.6013 | 0.9035 | 0.3001 |
| 0.0586 | 13.0 | 2847 | 0.1914 | 0.5471 | 0.7342 | 0.9402 | nan | 0.9758 | 0.7103 | 0.9814 | 0.2693 | 0.0 | 0.9290 | 0.6397 | 0.9150 | 0.2517 |
| 0.0531 | 14.0 | 3066 | 0.1747 | 0.5716 | 0.7689 | 0.9449 | nan | 0.9701 | 0.7945 | 0.9588 | 0.3522 | 0.0 | 0.9339 | 0.6815 | 0.9280 | 0.3147 |
| 0.0522 | 15.0 | 3285 | 0.1933 | 0.5591 | 0.7399 | 0.9454 | nan | 0.9810 | 0.7222 | 0.9744 | 0.2820 | 0.0 | 0.9351 | 0.6603 | 0.9355 | 0.2645 |
| 0.059 | 16.0 | 3504 | 0.1897 | 0.5691 | 0.7878 | 0.9384 | nan | 0.9499 | 0.8594 | 0.9809 | 0.3608 | 0.0 | 0.9252 | 0.6741 | 0.9159 | 0.3303 |
| 0.0503 | 17.0 | 3723 | 0.1895 | 0.5652 | 0.7795 | 0.9365 | nan | 0.9588 | 0.7866 | 0.9808 | 0.3917 | 0.0 | 0.9238 | 0.6508 | 0.9004 | 0.3511 |
| 0.0518 | 18.0 | 3942 | 0.2131 | 0.5533 | 0.7332 | 0.9402 | nan | 0.9807 | 0.6877 | 0.9645 | 0.2998 | 0.0 | 0.9294 | 0.6248 | 0.9334 | 0.2790 |
| 0.0439 | 19.0 | 4161 | 0.2168 | 0.5565 | 0.7411 | 0.9388 | nan | 0.9801 | 0.6828 | 0.9567 | 0.3448 | 0.0 | 0.9278 | 0.6194 | 0.9234 | 0.3121 |
| 0.0459 | 20.0 | 4380 | 0.2688 | 0.5266 | 0.7127 | 0.9266 | nan | 0.9824 | 0.5567 | 0.9841 | 0.3277 | 0.0 | 0.9149 | 0.5329 | 0.8866 | 0.2987 |
| 0.043 | 21.0 | 4599 | 0.2395 | 0.5542 | 0.7409 | 0.9369 | nan | 0.9821 | 0.6444 | 0.9745 | 0.3625 | 0.0 | 0.9258 | 0.5974 | 0.9228 | 0.3248 |
| 0.0436 | 22.0 | 4818 | 0.1790 | 0.5736 | 0.7750 | 0.9441 | nan | 0.9706 | 0.7783 | 0.9694 | 0.3819 | 0.0 | 0.9331 | 0.6772 | 0.9143 | 0.3433 |
| 0.0443 | 23.0 | 5037 | 0.1843 | 0.5683 | 0.7613 | 0.9442 | nan | 0.9756 | 0.7470 | 0.9716 | 0.3511 | 0.0 | 0.9335 | 0.6684 | 0.9177 | 0.3219 |
| 0.0402 | 24.0 | 5256 | 0.2048 | 0.5666 | 0.7535 | 0.9429 | nan | 0.9800 | 0.7089 | 0.9706 | 0.3544 | 0.0 | 0.9324 | 0.6457 | 0.9302 | 0.3246 |
| 0.0399 | 25.0 | 5475 | 0.2102 | 0.5651 | 0.7524 | 0.9430 | nan | 0.9830 | 0.6875 | 0.9754 | 0.3637 | 0.0 | 0.9327 | 0.6412 | 0.9231 | 0.3287 |
| 0.0404 | 26.0 | 5694 | 0.1993 | 0.5792 | 0.7815 | 0.9460 | nan | 0.9690 | 0.8035 | 0.9697 | 0.3837 | 0.0 | 0.9351 | 0.6876 | 0.9289 | 0.3443 |
| 0.0388 | 27.0 | 5913 | 0.2024 | 0.5681 | 0.7501 | 0.9470 | nan | 0.9821 | 0.7343 | 0.9605 | 0.3236 | 0.0 | 0.9370 | 0.6715 | 0.9322 | 0.3001 |
| 0.0369 | 28.0 | 6132 | 0.1830 | 0.5701 | 0.7553 | 0.9481 | nan | 0.9779 | 0.7698 | 0.9608 | 0.3126 | 0.0 | 0.9379 | 0.6871 | 0.9323 | 0.2931 |
| 0.0373 | 29.0 | 6351 | 0.2162 | 0.5682 | 0.7535 | 0.9438 | nan | 0.9828 | 0.7011 | 0.9639 | 0.3665 | 0.0 | 0.9335 | 0.6482 | 0.9239 | 0.3352 |
| 0.0348 | 30.0 | 6570 | 0.2126 | 0.5640 | 0.7479 | 0.9435 | nan | 0.9813 | 0.7097 | 0.9623 | 0.3384 | 0.0 | 0.9330 | 0.6537 | 0.9197 | 0.3135 |
| 0.0354 | 31.0 | 6789 | 0.2025 | 0.5626 | 0.7467 | 0.9469 | nan | 0.9795 | 0.7453 | 0.9725 | 0.2896 | 0.0 | 0.9368 | 0.6762 | 0.9285 | 0.2716 |
| 0.0344 | 32.0 | 7008 | 0.1973 | 0.5786 | 0.7739 | 0.9469 | nan | 0.9734 | 0.7828 | 0.9698 | 0.3695 | 0.0 | 0.9364 | 0.6853 | 0.9326 | 0.3389 |
| 0.0333 | 33.0 | 7227 | 0.2199 | 0.5722 | 0.7624 | 0.9438 | nan | 0.9817 | 0.7045 | 0.9696 | 0.3940 | 0.0 | 0.9334 | 0.6481 | 0.9287 | 0.3510 |
| 0.0345 | 34.0 | 7446 | 0.2052 | 0.5791 | 0.7724 | 0.9465 | nan | 0.9799 | 0.7347 | 0.9736 | 0.4015 | 0.0 | 0.9363 | 0.6698 | 0.9311 | 0.3582 |
| 0.0326 | 35.0 | 7665 | 0.2176 | 0.5758 | 0.7629 | 0.9462 | nan | 0.9835 | 0.7124 | 0.9689 | 0.3868 | 0.0 | 0.9362 | 0.6595 | 0.9345 | 0.3490 |
| 0.034 | 36.0 | 7884 | 0.2247 | 0.5717 | 0.7557 | 0.9453 | nan | 0.9841 | 0.7033 | 0.9661 | 0.3694 | 0.0 | 0.9352 | 0.6533 | 0.9331 | 0.3369 |
| 0.0324 | 37.0 | 8103 | 0.1957 | 0.5797 | 0.7736 | 0.9490 | nan | 0.9763 | 0.7801 | 0.9725 | 0.3657 | 0.0 | 0.9390 | 0.6963 | 0.9299 | 0.3333 |
| 0.0332 | 38.0 | 8322 | 0.1996 | 0.5770 | 0.7644 | 0.9478 | nan | 0.9826 | 0.7310 | 0.9696 | 0.3743 | 0.0 | 0.9379 | 0.6741 | 0.9336 | 0.3393 |
| 0.0332 | 39.0 | 8541 | 0.2129 | 0.5638 | 0.7423 | 0.9449 | nan | 0.9845 | 0.7021 | 0.9616 | 0.3212 | 0.0 | 0.9348 | 0.6514 | 0.9328 | 0.3001 |
| 0.03 | 40.0 | 8760 | 0.2283 | 0.5694 | 0.7539 | 0.9441 | nan | 0.9840 | 0.6931 | 0.9686 | 0.3699 | 0.0 | 0.9339 | 0.6464 | 0.9277 | 0.3387 |
| 0.0319 | 41.0 | 8979 | 0.2013 | 0.5741 | 0.7624 | 0.9471 | nan | 0.9804 | 0.7416 | 0.9670 | 0.3606 | 0.0 | 0.9370 | 0.6760 | 0.9277 | 0.3300 |
| 0.0361 | 42.0 | 9198 | 0.2094 | 0.5709 | 0.7568 | 0.9463 | nan | 0.9810 | 0.7317 | 0.9663 | 0.3483 | 0.0 | 0.9362 | 0.6689 | 0.9279 | 0.3216 |
| 0.0304 | 43.0 | 9417 | 0.2098 | 0.5731 | 0.7586 | 0.9468 | nan | 0.9821 | 0.7282 | 0.9666 | 0.3575 | 0.0 | 0.9368 | 0.6700 | 0.9295 | 0.3293 |
| 0.0303 | 44.0 | 9636 | 0.2155 | 0.5705 | 0.7554 | 0.9470 | nan | 0.9814 | 0.7329 | 0.9702 | 0.3370 | 0.0 | 0.9369 | 0.6718 | 0.9301 | 0.3137 |
| 0.03 | 45.0 | 9855 | 0.2183 | 0.5703 | 0.7541 | 0.9464 | nan | 0.9825 | 0.7229 | 0.9677 | 0.3435 | 0.0 | 0.9364 | 0.6657 | 0.9311 | 0.3181 |
| 0.0301 | 45.66 | 10000 | 0.2142 | 0.5724 | 0.7571 | 0.9468 | nan | 0.9826 | 0.7246 | 0.9679 | 0.3533 | 0.0 | 0.9368 | 0.6678 | 0.9310 | 0.3263 |
### Framework versions
- Transformers 4.31.0.dev0
- Pytorch 2.0.1+cpu
- Datasets 2.13.1
- Tokenizers 0.13.3
|
rhplus0831/maid-yuzu-v2 | rhplus0831 | 2024-02-01T02:08:39Z | 4 | 0 | transformers | [
"transformers",
"safetensors",
"mixtral",
"text-generation",
"mergekit",
"merge",
"base_model:smelborp/MixtralOrochi8x7B",
"base_model:merge:smelborp/MixtralOrochi8x7B",
"base_model:ycros/BagelMIsteryTour-v2-8x7B",
"base_model:merge:ycros/BagelMIsteryTour-v2-8x7B",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T01:41:19Z | ---
base_model:
- smelborp/MixtralOrochi8x7B
- ycros/BagelMIsteryTour-v2-8x7B
tags:
- mergekit
- merge
---
# maid-yuzu-v2
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
This model is an experiment to combine two models I liked.
## Merge Details
### Merge Method
This model was merged using the SLERP merge method.
### Models Merged
The following models were included in the merge:
* [smelborp/MixtralOrochi8x7B](https://huggingface.co/smelborp/MixtralOrochi8x7B)
* [ycros/BagelMIsteryTour-v2-8x7B](https://huggingface.co/ycros/BagelMIsteryTour-v2-8x7B)
### Configuration
The following YAML configuration was used to produce this model:
```yaml
base_model:
model:
path: smelborp/MixtralOrochi8x7B
dtype: bfloat16
merge_method: slerp
parameters:
t:
- value: 0.25
slices:
- sources:
- layer_range: [0, 32]
model:
model:
path: smelborp/MixtralOrochi8x7B
- layer_range: [0, 32]
model:
model:
path: ycros/BagelMIsteryTour-v2-8x7B
```
|
fuyu-quant/ibl-regression-ver4-branch | fuyu-quant | 2024-02-01T01:59:41Z | 1 | 0 | peft | [
"peft",
"region:us"
]
| null | 2024-01-31T16:08:50Z | ---
library_name: peft
---
## Training procedure
The following `bitsandbytes` quantization config was used during training:
- 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
### Framework versions
- PEFT 0.4.0
|
judith0/v1-ineClassification | judith0 | 2024-02-01T01:52:06Z | 196 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"swin",
"image-classification",
"generated_from_trainer",
"dataset:imagefolder",
"base_model:microsoft/swin-tiny-patch4-window7-224",
"base_model:finetune:microsoft/swin-tiny-patch4-window7-224",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| image-classification | 2024-02-01T01:50:32Z | ---
license: apache-2.0
base_model: microsoft/swin-tiny-patch4-window7-224
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
model-index:
- name: swin-tiny-patch4-window7-224-finetuned-eurosat
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: imagefolder
type: imagefolder
config: default
split: train
args: default
metrics:
- name: Accuracy
type: accuracy
value: 1.0
---
<!-- 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. -->
# swin-tiny-patch4-window7-224-finetuned-eurosat
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1430
- Accuracy: 1.0
## 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
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- 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 | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log | 0.89 | 6 | 0.5325 | 0.8936 |
| 0.833 | 1.93 | 13 | 0.1850 | 0.9787 |
| 0.833 | 2.67 | 18 | 0.1430 | 1.0 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
Castling/marian-finetuned-kde4-en-to-fr | Castling | 2024-02-01T01:44:06Z | 125 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"marian",
"text2text-generation",
"translation",
"generated_from_trainer",
"dataset:kde4",
"base_model:Helsinki-NLP/opus-mt-en-fr",
"base_model:finetune:Helsinki-NLP/opus-mt-en-fr",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
]
| translation | 2023-12-29T00:54:45Z | ---
license: apache-2.0
base_model: Helsinki-NLP/opus-mt-en-fr
tags:
- translation
- generated_from_trainer
datasets:
- kde4
metrics:
- bleu
model-index:
- name: marian-finetuned-kde4-en-to-fr
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: kde4
type: kde4
config: en-fr
split: train
args: en-fr
metrics:
- name: Bleu
type: bleu
value: 52.837727401681214
---
<!-- 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. -->
# marian-finetuned-kde4-en-to-fr
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-fr](https://huggingface.co/Helsinki-NLP/opus-mt-en-fr) on the kde4 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8556
- Bleu: 52.8377
## 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: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0
|
BachNgoH/a2c-PandaPushDense-v2 | BachNgoH | 2024-02-01T01:36:37Z | 2 | 0 | stable-baselines3 | [
"stable-baselines3",
"PandaPushDense-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"arxiv:2106.13687",
"model-index",
"region:us"
]
| reinforcement-learning | 2023-02-07T14:53:03Z | ---
library_name: stable-baselines3
tags:
- PandaPushDense-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: A2C
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: PandaPushDense-v2
type: PandaPushDense-v2
metrics:
- type: mean_reward
value: -7.94 +/- 4.62
name: mean_reward
verified: false
---
# **A2C** Agent playing **PandaPushDense-v2**
This is a trained model of a **A2C** agent playing **PandaPushDense-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
...
```
Panda Gym environments: [arxiv.org/abs/2106.13687](https://arxiv.org/abs/2106.13687) |
r3m3c3/english-to-kanji-c20000_2 | r3m3c3 | 2024-02-01T01:35:37Z | 29 | 0 | diffusers | [
"diffusers",
"safetensors",
"arxiv:1910.09700",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
]
| text-to-image | 2024-02-01T01:34:20Z | ---
library_name: diffusers
---
# 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. -->
This is the model card of a ๐งจ diffusers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
SurfaceData/llava-v1.6-vicuna-7b-processor | SurfaceData | 2024-02-01T01:33:22Z | 0 | 1 | transformers | [
"transformers",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| null | 2024-02-01T01:33:19Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
jbottarini/rl_workshop_1 | jbottarini | 2024-02-01T01:16:15Z | 0 | 0 | null | [
"Pixelcopter-PLE-v0",
"reinforce",
"reinforcement-learning",
"custom-implementation",
"deep-rl-class",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T01:14:37Z | ---
tags:
- Pixelcopter-PLE-v0
- reinforce
- reinforcement-learning
- custom-implementation
- deep-rl-class
model-index:
- name: rl_workshop_1
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Pixelcopter-PLE-v0
type: Pixelcopter-PLE-v0
metrics:
- type: mean_reward
value: 13.90 +/- 10.89
name: mean_reward
verified: false
---
# **Reinforce** Agent playing **Pixelcopter-PLE-v0**
This is a trained model of a **Reinforce** agent playing **Pixelcopter-PLE-v0** .
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
|
paisanx/Reinforce-Pixelcopter-PLE-v7 | paisanx | 2024-02-01T01:15:38Z | 0 | 0 | null | [
"Pixelcopter-PLE-v0",
"reinforce",
"reinforcement-learning",
"custom-implementation",
"deep-rl-class",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T01:15:31Z | ---
tags:
- Pixelcopter-PLE-v0
- reinforce
- reinforcement-learning
- custom-implementation
- deep-rl-class
model-index:
- name: Reinforce-Pixelcopter-PLE-v7
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Pixelcopter-PLE-v0
type: Pixelcopter-PLE-v0
metrics:
- type: mean_reward
value: -5.00 +/- 0.00
name: mean_reward
verified: false
---
# **Reinforce** Agent playing **Pixelcopter-PLE-v0**
This is a trained model of a **Reinforce** agent playing **Pixelcopter-PLE-v0** .
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
|
nanalysenko/panacea-test-2 | nanalysenko | 2024-02-01T01:02:16Z | 44 | 0 | sentence-transformers | [
"sentence-transformers",
"safetensors",
"roberta",
"feature-extraction",
"sentence-similarity",
"transformers",
"autotrain_compatible",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
]
| sentence-similarity | 2024-02-01T00:56:03Z | ---
library_name: sentence-transformers
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
---
# {MODEL_NAME}
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
<!--- Describe your model here -->
## Usage (Sentence-Transformers)
Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
```
pip install -U sentence-transformers
```
Then you can use the model like this:
```python
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('{MODEL_NAME}')
embeddings = model.encode(sentences)
print(embeddings)
```
## Usage (HuggingFace Transformers)
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
```python
from transformers import AutoTokenizer, AutoModel
import torch
#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0] #First element of model_output contains all token embeddings
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']
# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('{MODEL_NAME}')
model = AutoModel.from_pretrained('{MODEL_NAME}')
# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
# Compute token embeddings
with torch.no_grad():
model_output = model(**encoded_input)
# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
print("Sentence embeddings:")
print(sentence_embeddings)
```
## Evaluation Results
<!--- Describe how your model was evaluated -->
For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name={MODEL_NAME})
## Training
The model was trained with the parameters:
**DataLoader**:
`torch.utils.data.dataloader.DataLoader` of length 75 with parameters:
```
{'batch_size': 16, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
```
**Loss**:
`sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss` with parameters:
```
{'scale': 20.0, 'similarity_fct': 'cos_sim'}
```
Parameters of the fit()-Method:
```
{
"epochs": 10,
"evaluation_steps": 0,
"evaluator": "NoneType",
"max_grad_norm": 1,
"optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
"optimizer_params": {
"lr": 2e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 75,
"weight_decay": 0.01
}
```
## Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: RobertaModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False})
)
```
## Citing & Authors
<!--- Describe where people can find more information --> |
asun17904/anliR1-t5-base-kd | asun17904 | 2024-02-01T00:35:01Z | 0 | 0 | pytorch | [
"pytorch",
"en",
"license:mit",
"region:us"
]
| null | 2024-01-31T05:17:44Z | ---
language: en
license: mit
library_name: pytorch
---
# Knowledge Continuity Regularized Network
Dataset: ANLI
Round: None
Trainer Hyperparameters:
- `lr` = 5e-05
- `per_device_batch_size` = 8
- `gradient_accumulation_steps` = 2
- `weight_decay` = 0.0
- `seed` = 42
Regularization Hyperparameters
- `numerical stability denominator constant` = 0.01
- `lambda` = 0.0001
- `alpha` = 2.0
- `beta` = 2.0
Extended Logs:
|eval_loss|eval_accuracy|epoch|
|--|--|--|
|34.239|0.403|1.0|
|35.481|0.395|2.0|
|35.944|0.407|3.0|
|35.125|0.438|4.0|
|34.990|0.444|5.0|
|35.375|0.425|6.0|
|34.941|0.449|7.0|
|34.913|0.447|8.0|
|34.559|0.461|9.0|
|34.499|0.464|10.0|
|34.479|0.462|11.0|
|34.368|0.461|12.0|
|34.369|0.464|13.0|
|34.496|0.466|14.0|
|34.358|0.468|15.0|
|34.275|0.469|16.0|
|34.063|0.477|17.0|
|33.969|0.483|18.0|
|33.991|0.478|19.0|
**Test Accuracy: 0.482** |
Seher99/mistral-7b-dollyy | Seher99 | 2024-02-01T00:31:52Z | 0 | 0 | transformers | [
"transformers",
"safetensors",
"arxiv:1910.09700",
"endpoints_compatible",
"region:us"
]
| null | 2024-01-30T00:29:00Z | ---
library_name: transformers
tags: []
---
# 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. -->
This is the model card of a ๐ค transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
XiaQiang/q-FrozenLake-v1-4x4-noSlippery | XiaQiang | 2024-02-01T00:24:33Z | 0 | 0 | null | [
"FrozenLake-v1-4x4-no_slippery",
"q-learning",
"reinforcement-learning",
"custom-implementation",
"model-index",
"region:us"
]
| reinforcement-learning | 2024-02-01T00:24:31Z | ---
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="XiaQiang/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"])
```
|
asun17904/glue-rte-bert-base-uncased | asun17904 | 2024-02-01T00:20:09Z | 0 | 0 | pytorch | [
"pytorch",
"en",
"license:mit",
"region:us"
]
| null | 2024-02-01T00:12:15Z | ---
language: en
license: mit
library_name: pytorch
---
# Plainly Optimized Network
Dataset: GLUE
Trainer Hyperparameters:
- `lr` = 5e-05
- `per_device_batch_size` = 32
- `gradient_accumulation_steps` = 1
- `weight_decay` = 0.0
- `seed` = 42
|eval_loss|eval_accuracy|epoch|
|--|--|--|
|21.354|0.531|1.0|
|20.627|0.621|2.0|
|20.929|0.610|3.0|
|21.703|0.585|4.0|
|21.829|0.581|5.0|
|21.040|0.635|6.0|
|21.964|0.585|7.0|
|21.910|0.581|8.0|
|22.344|0.581|9.0|
|
AIFT/AIFT-instruct-v1.3-42dot_LLM-SFT-1.3B | AIFT | 2024-02-01T00:19:21Z | 150 | 0 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"license:cc-by-sa-4.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T00:17:32Z | ---
license: cc-by-sa-4.0
---
<h1>AIFT-instruct-42dot_LLM-SFT-1.3B</h1>
<b><ํ์ต ๋ฐ์ดํฐ ๊ตฌ์ถ></b>
<br>
kyujinpy ๋์ด ๊ณต๊ฐํ์ KOR-OpenOrca-Platypus ๋ฐ์ดํฐ๋ฅผ ์ผ๋ถ ์ญ์ (์ํ๋ง) ๋ฐ ์ ์ ์์
์งํํ์ฌ ํ์ฉ.
๊ทธ ์ดํ ํด๋น ๋ฐ์ดํฐ๋ค์ ๋ณด๋ฉฐ ๊ด๋ จ ํ์คํฌ๋ฅผ ์ถ์ถํ์๊ณ ์ด๋ฅผ ๊ธฐ๋ฐ์ผ๋ก
ํด๋น ํ์คํฌ์ ๋ง์ถฐ์ NLP ๊ด๋ จ ์คํ์์ค ๋ฐ์ดํฐ๋ฅผ ํ์ฉํ์ฌ ํ์ต๋ฐ์ดํฐ๋ฅผ ์์ฒด์ ์ผ๋ก
์ญ์ฌ, ๊ณผํ, ์ํ, ๊ธฐ๊ณ๋
ํด, ๋ฆฌ๋ทฐ ๋ถ์ ๋ฌธ์ ๋ฅผ gpt๋ฅผ ํตํด์ ๊ตฌ์ถํ์๊ณ ,
aihub ์ผ๋ฐ์์ ๋ฐ ๊ธฐ๊ณ๋
ํด ๋ฐ์ดํฐ๋ฅผ ํ์ฉํ์ฌ ์ถ๊ฐ๋ก ํ์ต ๋ฐ์ดํฐ๋ฅผ ๊ตฌ์ถ(ํํ์ ๊ด๋ จ, ๊ธฐ๊ณ๋
ํด ๊ด๋ จ ๋ฐ ์์ฝ)
๊ฐ์ข
๋ธ๋ก๊ทธ์์ ์ญ์ฌ ๋ฐ ์์ ํด์ฆ๋ฅผ ์ฌ๋์ด ์ง์ ํ์ต๋ฐ์ดํฐ ํํ๋ก ๋ณ๊ฒฝ
AI2AI Challenge ๋ฐ์ดํฐ ํํ๋ฅผ ๋ณด๊ณ gpt๋ฅผ ํตํด ์ด๋ฑ ์์ค์ ๊ณผํ ์ํ ๋ฌธ์ ์ ํ์ ์ ์ 500๋ฌธ์
์์ด ๋ฒ์ญ ๋ฐ์ดํฐ ์ํ/ํ์ ๋ฐ์ดํฐ ํ์ต ๋ฐ์ดํฐ๋ก ํ์ฉ ์งํ
์ด ๋ฐ์ดํฐ 4๋ง๊ฐ ์ ๋ ์ฌ์ฉํ์์ต๋๋ค.
<br>
<br>
+ TruthfulQA ๊ด๋ จ ๋ฌธ์ ์ถ๊ฐ๋ฅผ ์งํํ์์ต๋๋ค.(์์ค ๊ด๋ จ ์ฐธ๊ฑฐ์ง ๋ฌธ์ )
+ ๊ธฐ๊ณ๋
ํด ๊ด๋ จ ํ์ต ๋ฐ์ดํฐ๋ฅผ ChatGPT๋ฅผ ํตํด์ ๋ต๋ณ์ ์ป์ด ํ์ต
+ ๋ฌธ๋ฒ๊ด๋ จ ํ์ต ๋ฐ์ดํฐ
<br>
###ํ์ต ๋ฐ์ดํฐ ํ์ผ์ ๋น๊ณต๊ฐ์
๋๋ค.
<br>
<๋ชจ๋ธ>
<br>
42dot์์ ๊ณต๊ฐํ 42dot_LLM-SFT-1.3B์ ๋ฒ ์ด์ค ๋ชจ๋ธ๋ก ํ์ฌ ํ์ต ์งํํ์์ต๋๋ค.
<br>
<br>
<br>
<b><ํ์ต></b>
<br>
ํ์ต์ LoRA๋ฅผ ์ฌ์ฉํ์ฌ A100 40G *2์์ ํ์ต์ ์งํํ์์ต๋๋ค.
|
Zintoulou/finetuningnewmodule1 | Zintoulou | 2024-02-01T00:14:58Z | 1 | 0 | peft | [
"peft",
"tensorboard",
"safetensors",
"generated_from_trainer",
"base_model:codellama/CodeLlama-7b-Instruct-hf",
"base_model:adapter:codellama/CodeLlama-7b-Instruct-hf",
"license:llama2",
"region:us"
]
| null | 2024-01-31T23:29:48Z | ---
license: llama2
library_name: peft
tags:
- generated_from_trainer
base_model: codellama/CodeLlama-7b-Instruct-hf
model-index:
- name: finetuningnewmodule1
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. -->
# finetuningnewmodule1
This model is a fine-tuned version of [codellama/CodeLlama-7b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-7b-Instruct-hf) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.9468
## 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: 0.001
- 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: 8
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 2.688 | 1.0 | 1 | 2.7045 |
| 2.2248 | 2.0 | 2 | 2.0812 |
| 1.6447 | 3.0 | 3 | 1.7351 |
| 1.2926 | 4.0 | 4 | 1.2896 |
| 0.8111 | 5.0 | 5 | 1.0243 |
| 0.4457 | 6.0 | 6 | 0.9231 |
| 0.2562 | 7.0 | 7 | 0.9348 |
| 0.1901 | 8.0 | 8 | 0.9468 |
### Framework versions
- Transformers 4.36.0
- Pytorch 2.0.1
- Datasets 2.16.1
- Tokenizers 0.15.1
## Training procedure
### Framework versions
- PEFT 0.6.0
|
son-of-man/HoloViolet-7B-test3 | son-of-man | 2024-02-01T00:04:15Z | 4 | 0 | transformers | [
"transformers",
"safetensors",
"mistral",
"text-generation",
"merge",
"mergekit",
"lazymergekit",
"GreenNode/GreenNode-mini-7B-multilingual-v1olet",
"KoboldAI/Mistral-7B-Holodeck-1",
"FPHam/Sydney_Pirate_Mistral_7b",
"base_model:FPHam/Sydney_Pirate_Mistral_7b",
"base_model:merge:FPHam/Sydney_Pirate_Mistral_7b",
"base_model:GreenNode/GreenNode-mini-7B-multilingual-v1olet",
"base_model:merge:GreenNode/GreenNode-mini-7B-multilingual-v1olet",
"base_model:KoboldAI/Mistral-7B-Holodeck-1",
"base_model:merge:KoboldAI/Mistral-7B-Holodeck-1",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
]
| text-generation | 2024-02-01T00:00:22Z | ---
tags:
- merge
- mergekit
- lazymergekit
- GreenNode/GreenNode-mini-7B-multilingual-v1olet
- KoboldAI/Mistral-7B-Holodeck-1
- FPHam/Sydney_Pirate_Mistral_7b
base_model:
- GreenNode/GreenNode-mini-7B-multilingual-v1olet
- KoboldAI/Mistral-7B-Holodeck-1
- FPHam/Sydney_Pirate_Mistral_7b
---
# HoloViolet-7B-test3
HoloViolet-7B-test3 is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [GreenNode/GreenNode-mini-7B-multilingual-v1olet](https://huggingface.co/GreenNode/GreenNode-mini-7B-multilingual-v1olet)
* [KoboldAI/Mistral-7B-Holodeck-1](https://huggingface.co/KoboldAI/Mistral-7B-Holodeck-1)
* [FPHam/Sydney_Pirate_Mistral_7b](https://huggingface.co/FPHam/Sydney_Pirate_Mistral_7b)
## ๐งฉ Configuration
```yaml
merge_method: task_arithmetic
base_model: mistralai/Mistral-7B-v0.1
models:
- model: mistralai/Mistral-7B-v0.1
- model: GreenNode/GreenNode-mini-7B-multilingual-v1olet
parameters:
weight: 0.69
- model: KoboldAI/Mistral-7B-Holodeck-1
parameters:
weight: 0.42
- model: FPHam/Sydney_Pirate_Mistral_7b
parameters:
weight: 0.21
dtype: bfloat16
```
## ๐ป Usage
```python
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "son-of-man/HoloViolet-7B-test3"
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"])
``` |
r3m3c3/english-to-kanji-c14500_2 | r3m3c3 | 2024-01-31T23:50:26Z | 29 | 0 | diffusers | [
"diffusers",
"safetensors",
"arxiv:1910.09700",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
]
| text-to-image | 2024-01-31T23:49:16Z | ---
library_name: diffusers
---
# 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. -->
This is the model card of a ๐งจ diffusers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
r3m3c3/english-to-kanji-c8000_2 | r3m3c3 | 2024-01-31T23:47:14Z | 29 | 0 | diffusers | [
"diffusers",
"safetensors",
"arxiv:1910.09700",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
]
| text-to-image | 2024-01-31T23:46:06Z | ---
library_name: diffusers
---
# 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. -->
This is the model card of a ๐งจ diffusers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
r3m3c3/english-to-kanji-c7000_2 | r3m3c3 | 2024-01-31T23:45:36Z | 29 | 0 | diffusers | [
"diffusers",
"safetensors",
"arxiv:1910.09700",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
]
| text-to-image | 2024-01-31T23:44:25Z | ---
library_name: diffusers
---
# 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. -->
This is the model card of a ๐งจ diffusers model that has been pushed on the Hub. This model card has been automatically generated.
- **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]
|
rasyosef/gpt2-oscar-amharic-tokenizer | rasyosef | 2024-01-31T23:37:13Z | 0 | 0 | transformers | [
"transformers",
"am",
"dataset:oscar",
"license:mit",
"endpoints_compatible",
"region:us"
]
| null | 2024-01-31T23:05:31Z | ---
license: mit
datasets:
- oscar
language:
- am
library_name: transformers
---
# Amharic BPE Tokenizer
This repo contains a **Byte-Pair Encoding** tokenizer trained on the **Amharic** subset of the [oscar](https://huggingface.co/datasets/oscar) dataset. It's the same as the GPT-2 tokenizer but trained from scratch on an amharic dataset with a vocabulary size of `24000`.
# How to use
You can load the tokenizer from huggingface hub as follows.
```python
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("rasyosef/gpt2-oscar-amharic-tokenizer")
tokenizer("แ แฃแญแ แซแแจ แจแแแ แฒแฌแต แฅแฝแแแแแข")
``` |
Subsets and Splits