End of training
Browse files- README.md +106 -0
- adapter_config.json +32 -0
- adapter_model.safetensors +3 -0
- all_results.json +14 -0
- config.json +27 -0
- eval_results.json +8 -0
- special_tokens_map.json +28 -0
- tokenizer.json +0 -0
- tokenizer_config.json +62 -0
- train_results.json +9 -0
- trainer_state.json +630 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- text-generation
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- sft
|
| 9 |
+
base_model: mistralai/Mistral-7B-Instruct-v0.2
|
| 10 |
+
datasets:
|
| 11 |
+
- generator
|
| 12 |
+
model-index:
|
| 13 |
+
- name: mistral_7b_instruct_v2_constitutional_rf_v2
|
| 14 |
+
results: []
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 18 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 19 |
+
|
| 20 |
+
# mistral_7b_instruct_v2_constitutional_rf_v2
|
| 21 |
+
|
| 22 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the generator dataset.
|
| 23 |
+
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 1.1829
|
| 25 |
+
|
| 26 |
+
## Model description
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Intended uses & limitations
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training and evaluation data
|
| 35 |
+
|
| 36 |
+
More information needed
|
| 37 |
+
|
| 38 |
+
## Training procedure
|
| 39 |
+
|
| 40 |
+
### Training hyperparameters
|
| 41 |
+
|
| 42 |
+
The following hyperparameters were used during training:
|
| 43 |
+
- learning_rate: 0.0002
|
| 44 |
+
- train_batch_size: 1
|
| 45 |
+
- eval_batch_size: 8
|
| 46 |
+
- seed: 42
|
| 47 |
+
- gradient_accumulation_steps: 4
|
| 48 |
+
- total_train_batch_size: 4
|
| 49 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 50 |
+
- lr_scheduler_type: linear
|
| 51 |
+
- lr_scheduler_warmup_steps: 1
|
| 52 |
+
- training_steps: 1000
|
| 53 |
+
|
| 54 |
+
### Training results
|
| 55 |
+
|
| 56 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 57 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 58 |
+
| 0.9391 | 0.1479 | 25 | 0.6647 |
|
| 59 |
+
| 0.6132 | 0.2959 | 50 | 0.6131 |
|
| 60 |
+
| 0.6037 | 0.4438 | 75 | 0.6059 |
|
| 61 |
+
| 0.5924 | 0.5917 | 100 | 0.5997 |
|
| 62 |
+
| 0.5962 | 0.7396 | 125 | 0.5944 |
|
| 63 |
+
| 0.6017 | 0.8876 | 150 | 0.5940 |
|
| 64 |
+
| 0.546 | 1.0355 | 175 | 0.6261 |
|
| 65 |
+
| 0.4233 | 1.1834 | 200 | 0.6179 |
|
| 66 |
+
| 0.4295 | 1.3314 | 225 | 0.6140 |
|
| 67 |
+
| 0.4165 | 1.4793 | 250 | 0.6192 |
|
| 68 |
+
| 0.428 | 1.6272 | 275 | 0.6221 |
|
| 69 |
+
| 0.4256 | 1.7751 | 300 | 0.6257 |
|
| 70 |
+
| 0.4254 | 1.9231 | 325 | 0.6235 |
|
| 71 |
+
| 0.3561 | 2.0710 | 350 | 0.6811 |
|
| 72 |
+
| 0.274 | 2.2189 | 375 | 0.7050 |
|
| 73 |
+
| 0.2681 | 2.3669 | 400 | 0.7119 |
|
| 74 |
+
| 0.2792 | 2.5148 | 425 | 0.7170 |
|
| 75 |
+
| 0.2738 | 2.6627 | 450 | 0.7101 |
|
| 76 |
+
| 0.2787 | 2.8107 | 475 | 0.7120 |
|
| 77 |
+
| 0.2794 | 2.9586 | 500 | 0.7014 |
|
| 78 |
+
| 0.1912 | 3.1065 | 525 | 0.8123 |
|
| 79 |
+
| 0.1592 | 3.2544 | 550 | 0.8475 |
|
| 80 |
+
| 0.1614 | 3.4024 | 575 | 0.8444 |
|
| 81 |
+
| 0.1628 | 3.5503 | 600 | 0.8430 |
|
| 82 |
+
| 0.1599 | 3.6982 | 625 | 0.8433 |
|
| 83 |
+
| 0.1667 | 3.8462 | 650 | 0.8468 |
|
| 84 |
+
| 0.1643 | 3.9941 | 675 | 0.8587 |
|
| 85 |
+
| 0.0958 | 4.1420 | 700 | 0.9951 |
|
| 86 |
+
| 0.0872 | 4.2899 | 725 | 1.0223 |
|
| 87 |
+
| 0.0892 | 4.4379 | 750 | 1.0246 |
|
| 88 |
+
| 0.0892 | 4.5858 | 775 | 1.0379 |
|
| 89 |
+
| 0.086 | 4.7337 | 800 | 1.0277 |
|
| 90 |
+
| 0.0871 | 4.8817 | 825 | 1.0588 |
|
| 91 |
+
| 0.0808 | 5.0296 | 850 | 1.0811 |
|
| 92 |
+
| 0.0512 | 5.1775 | 875 | 1.1683 |
|
| 93 |
+
| 0.0512 | 5.3254 | 900 | 1.1615 |
|
| 94 |
+
| 0.051 | 5.4734 | 925 | 1.1937 |
|
| 95 |
+
| 0.0493 | 5.6213 | 950 | 1.1847 |
|
| 96 |
+
| 0.0496 | 5.7692 | 975 | 1.1814 |
|
| 97 |
+
| 0.0469 | 5.9172 | 1000 | 1.1829 |
|
| 98 |
+
|
| 99 |
+
|
| 100 |
+
### Framework versions
|
| 101 |
+
|
| 102 |
+
- PEFT 0.10.0
|
| 103 |
+
- Transformers 4.40.0
|
| 104 |
+
- Pytorch 2.2.2
|
| 105 |
+
- Datasets 2.19.0
|
| 106 |
+
- Tokenizers 0.19.1
|
adapter_config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "mistralai/Mistral-7B-Instruct-v0.2",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layer_replication": null,
|
| 10 |
+
"layers_pattern": null,
|
| 11 |
+
"layers_to_transform": null,
|
| 12 |
+
"loftq_config": {},
|
| 13 |
+
"lora_alpha": 32,
|
| 14 |
+
"lora_dropout": 0.1,
|
| 15 |
+
"megatron_config": null,
|
| 16 |
+
"megatron_core": "megatron.core",
|
| 17 |
+
"modules_to_save": null,
|
| 18 |
+
"peft_type": "LORA",
|
| 19 |
+
"r": 32,
|
| 20 |
+
"rank_pattern": {},
|
| 21 |
+
"revision": null,
|
| 22 |
+
"target_modules": [
|
| 23 |
+
"k_proj",
|
| 24 |
+
"v_proj",
|
| 25 |
+
"q_proj",
|
| 26 |
+
"o_proj",
|
| 27 |
+
"gate_proj"
|
| 28 |
+
],
|
| 29 |
+
"task_type": "CAUSAL_LM",
|
| 30 |
+
"use_dora": false,
|
| 31 |
+
"use_rslora": false
|
| 32 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b6659cbb0f6159dfdcf91f0f74547a52dea221d2f7afd63f6b023c296272123
|
| 3 |
+
size 616639296
|
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.9171597633136095,
|
| 3 |
+
"eval_loss": 1.1828829050064087,
|
| 4 |
+
"eval_runtime": 5.5014,
|
| 5 |
+
"eval_samples": 169,
|
| 6 |
+
"eval_samples_per_second": 16.178,
|
| 7 |
+
"eval_steps_per_second": 2.181,
|
| 8 |
+
"total_flos": 1.75885655212032e+17,
|
| 9 |
+
"train_loss": 0.27826149678230283,
|
| 10 |
+
"train_runtime": 1217.2703,
|
| 11 |
+
"train_samples": 1346,
|
| 12 |
+
"train_samples_per_second": 3.286,
|
| 13 |
+
"train_steps_per_second": 0.822
|
| 14 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "mistralai/Mistral-7B-Instruct-v0.2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MistralForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 32000,
|
| 8 |
+
"eos_token_id": 32001,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 4096,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 14336,
|
| 13 |
+
"max_position_embeddings": 32768,
|
| 14 |
+
"model_type": "mistral",
|
| 15 |
+
"num_attention_heads": 32,
|
| 16 |
+
"num_hidden_layers": 32,
|
| 17 |
+
"num_key_value_heads": 8,
|
| 18 |
+
"pad_token_id": 32001,
|
| 19 |
+
"rms_norm_eps": 1e-05,
|
| 20 |
+
"rope_theta": 1000000.0,
|
| 21 |
+
"sliding_window": null,
|
| 22 |
+
"tie_word_embeddings": false,
|
| 23 |
+
"torch_dtype": "float16",
|
| 24 |
+
"transformers_version": "4.40.0",
|
| 25 |
+
"use_cache": true,
|
| 26 |
+
"vocab_size": 32002
|
| 27 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.9171597633136095,
|
| 3 |
+
"eval_loss": 1.1828829050064087,
|
| 4 |
+
"eval_runtime": 5.5014,
|
| 5 |
+
"eval_samples": 169,
|
| 6 |
+
"eval_samples_per_second": 16.178,
|
| 7 |
+
"eval_steps_per_second": 2.181
|
| 8 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
{
|
| 4 |
+
"content": "<|im_start|>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"content": "<|im_end|>",
|
| 12 |
+
"lstrip": false,
|
| 13 |
+
"normalized": false,
|
| 14 |
+
"rstrip": false,
|
| 15 |
+
"single_word": false
|
| 16 |
+
}
|
| 17 |
+
],
|
| 18 |
+
"bos_token": "<|im_start|>",
|
| 19 |
+
"eos_token": "<|im_end|>",
|
| 20 |
+
"pad_token": "<|im_end|>",
|
| 21 |
+
"unk_token": {
|
| 22 |
+
"content": "<unk>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false
|
| 27 |
+
}
|
| 28 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<unk>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"32000": {
|
| 30 |
+
"content": "<|im_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"32001": {
|
| 38 |
+
"content": "<|im_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
}
|
| 45 |
+
},
|
| 46 |
+
"additional_special_tokens": [
|
| 47 |
+
"<|im_start|>",
|
| 48 |
+
"<|im_end|>"
|
| 49 |
+
],
|
| 50 |
+
"bos_token": "<|im_start|>",
|
| 51 |
+
"chat_template": "{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
|
| 52 |
+
"clean_up_tokenization_spaces": false,
|
| 53 |
+
"eos_token": "<|im_end|>",
|
| 54 |
+
"legacy": true,
|
| 55 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 56 |
+
"pad_token": "<|im_end|>",
|
| 57 |
+
"sp_model_kwargs": {},
|
| 58 |
+
"spaces_between_special_tokens": false,
|
| 59 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 60 |
+
"unk_token": "<unk>",
|
| 61 |
+
"use_default_system_prompt": false
|
| 62 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.9171597633136095,
|
| 3 |
+
"total_flos": 1.75885655212032e+17,
|
| 4 |
+
"train_loss": 0.27826149678230283,
|
| 5 |
+
"train_runtime": 1217.2703,
|
| 6 |
+
"train_samples": 1346,
|
| 7 |
+
"train_samples_per_second": 3.286,
|
| 8 |
+
"train_steps_per_second": 0.822
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,630 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.9171597633136095,
|
| 5 |
+
"eval_steps": 25,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.14792899408284024,
|
| 13 |
+
"grad_norm": 0.83642578125,
|
| 14 |
+
"learning_rate": 0.0001951951951951952,
|
| 15 |
+
"loss": 0.9391,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.14792899408284024,
|
| 20 |
+
"eval_loss": 0.6647412776947021,
|
| 21 |
+
"eval_runtime": 5.4263,
|
| 22 |
+
"eval_samples_per_second": 16.402,
|
| 23 |
+
"eval_steps_per_second": 2.211,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2958579881656805,
|
| 28 |
+
"grad_norm": 0.64208984375,
|
| 29 |
+
"learning_rate": 0.0001901901901901902,
|
| 30 |
+
"loss": 0.6132,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.2958579881656805,
|
| 35 |
+
"eval_loss": 0.6130561232566833,
|
| 36 |
+
"eval_runtime": 5.4068,
|
| 37 |
+
"eval_samples_per_second": 16.461,
|
| 38 |
+
"eval_steps_per_second": 2.219,
|
| 39 |
+
"step": 50
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.4437869822485207,
|
| 43 |
+
"grad_norm": 0.7255859375,
|
| 44 |
+
"learning_rate": 0.0001851851851851852,
|
| 45 |
+
"loss": 0.6037,
|
| 46 |
+
"step": 75
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.4437869822485207,
|
| 50 |
+
"eval_loss": 0.605912983417511,
|
| 51 |
+
"eval_runtime": 5.4769,
|
| 52 |
+
"eval_samples_per_second": 16.25,
|
| 53 |
+
"eval_steps_per_second": 2.191,
|
| 54 |
+
"step": 75
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.591715976331361,
|
| 58 |
+
"grad_norm": 0.6396484375,
|
| 59 |
+
"learning_rate": 0.00018018018018018018,
|
| 60 |
+
"loss": 0.5924,
|
| 61 |
+
"step": 100
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.591715976331361,
|
| 65 |
+
"eval_loss": 0.5997055768966675,
|
| 66 |
+
"eval_runtime": 5.4232,
|
| 67 |
+
"eval_samples_per_second": 16.411,
|
| 68 |
+
"eval_steps_per_second": 2.213,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.7396449704142012,
|
| 73 |
+
"grad_norm": 0.6826171875,
|
| 74 |
+
"learning_rate": 0.0001751751751751752,
|
| 75 |
+
"loss": 0.5962,
|
| 76 |
+
"step": 125
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.7396449704142012,
|
| 80 |
+
"eval_loss": 0.5943592190742493,
|
| 81 |
+
"eval_runtime": 5.4562,
|
| 82 |
+
"eval_samples_per_second": 16.312,
|
| 83 |
+
"eval_steps_per_second": 2.199,
|
| 84 |
+
"step": 125
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.8875739644970414,
|
| 88 |
+
"grad_norm": 0.65771484375,
|
| 89 |
+
"learning_rate": 0.0001701701701701702,
|
| 90 |
+
"loss": 0.6017,
|
| 91 |
+
"step": 150
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.8875739644970414,
|
| 95 |
+
"eval_loss": 0.5939959287643433,
|
| 96 |
+
"eval_runtime": 5.4411,
|
| 97 |
+
"eval_samples_per_second": 16.357,
|
| 98 |
+
"eval_steps_per_second": 2.205,
|
| 99 |
+
"step": 150
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 1.0355029585798816,
|
| 103 |
+
"grad_norm": 0.5908203125,
|
| 104 |
+
"learning_rate": 0.00016516516516516518,
|
| 105 |
+
"loss": 0.546,
|
| 106 |
+
"step": 175
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 1.0355029585798816,
|
| 110 |
+
"eval_loss": 0.6260668039321899,
|
| 111 |
+
"eval_runtime": 5.4895,
|
| 112 |
+
"eval_samples_per_second": 16.213,
|
| 113 |
+
"eval_steps_per_second": 2.186,
|
| 114 |
+
"step": 175
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.183431952662722,
|
| 118 |
+
"grad_norm": 0.638671875,
|
| 119 |
+
"learning_rate": 0.00016016016016016018,
|
| 120 |
+
"loss": 0.4233,
|
| 121 |
+
"step": 200
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.183431952662722,
|
| 125 |
+
"eval_loss": 0.6178574562072754,
|
| 126 |
+
"eval_runtime": 5.5065,
|
| 127 |
+
"eval_samples_per_second": 16.163,
|
| 128 |
+
"eval_steps_per_second": 2.179,
|
| 129 |
+
"step": 200
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.331360946745562,
|
| 133 |
+
"grad_norm": 0.70947265625,
|
| 134 |
+
"learning_rate": 0.00015515515515515516,
|
| 135 |
+
"loss": 0.4295,
|
| 136 |
+
"step": 225
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.331360946745562,
|
| 140 |
+
"eval_loss": 0.6140432953834534,
|
| 141 |
+
"eval_runtime": 5.468,
|
| 142 |
+
"eval_samples_per_second": 16.276,
|
| 143 |
+
"eval_steps_per_second": 2.195,
|
| 144 |
+
"step": 225
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.4792899408284024,
|
| 148 |
+
"grad_norm": 0.685546875,
|
| 149 |
+
"learning_rate": 0.00015015015015015014,
|
| 150 |
+
"loss": 0.4165,
|
| 151 |
+
"step": 250
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.4792899408284024,
|
| 155 |
+
"eval_loss": 0.6192346811294556,
|
| 156 |
+
"eval_runtime": 5.4874,
|
| 157 |
+
"eval_samples_per_second": 16.219,
|
| 158 |
+
"eval_steps_per_second": 2.187,
|
| 159 |
+
"step": 250
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.6272189349112427,
|
| 163 |
+
"grad_norm": 0.82958984375,
|
| 164 |
+
"learning_rate": 0.00014514514514514515,
|
| 165 |
+
"loss": 0.428,
|
| 166 |
+
"step": 275
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 1.6272189349112427,
|
| 170 |
+
"eval_loss": 0.6221323609352112,
|
| 171 |
+
"eval_runtime": 5.4947,
|
| 172 |
+
"eval_samples_per_second": 16.197,
|
| 173 |
+
"eval_steps_per_second": 2.184,
|
| 174 |
+
"step": 275
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.7751479289940828,
|
| 178 |
+
"grad_norm": 0.6923828125,
|
| 179 |
+
"learning_rate": 0.00014014014014014013,
|
| 180 |
+
"loss": 0.4256,
|
| 181 |
+
"step": 300
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.7751479289940828,
|
| 185 |
+
"eval_loss": 0.6257030367851257,
|
| 186 |
+
"eval_runtime": 5.4444,
|
| 187 |
+
"eval_samples_per_second": 16.347,
|
| 188 |
+
"eval_steps_per_second": 2.204,
|
| 189 |
+
"step": 300
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.9230769230769231,
|
| 193 |
+
"grad_norm": 0.70703125,
|
| 194 |
+
"learning_rate": 0.00013513513513513514,
|
| 195 |
+
"loss": 0.4254,
|
| 196 |
+
"step": 325
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.9230769230769231,
|
| 200 |
+
"eval_loss": 0.6234580874443054,
|
| 201 |
+
"eval_runtime": 5.4501,
|
| 202 |
+
"eval_samples_per_second": 16.33,
|
| 203 |
+
"eval_steps_per_second": 2.202,
|
| 204 |
+
"step": 325
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 2.0710059171597632,
|
| 208 |
+
"grad_norm": 0.7587890625,
|
| 209 |
+
"learning_rate": 0.00013013013013013014,
|
| 210 |
+
"loss": 0.3561,
|
| 211 |
+
"step": 350
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 2.0710059171597632,
|
| 215 |
+
"eval_loss": 0.6810835599899292,
|
| 216 |
+
"eval_runtime": 5.4991,
|
| 217 |
+
"eval_samples_per_second": 16.184,
|
| 218 |
+
"eval_steps_per_second": 2.182,
|
| 219 |
+
"step": 350
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 2.2189349112426036,
|
| 223 |
+
"grad_norm": 0.72998046875,
|
| 224 |
+
"learning_rate": 0.00012512512512512512,
|
| 225 |
+
"loss": 0.274,
|
| 226 |
+
"step": 375
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 2.2189349112426036,
|
| 230 |
+
"eval_loss": 0.7050045132637024,
|
| 231 |
+
"eval_runtime": 5.4732,
|
| 232 |
+
"eval_samples_per_second": 16.261,
|
| 233 |
+
"eval_steps_per_second": 2.193,
|
| 234 |
+
"step": 375
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.366863905325444,
|
| 238 |
+
"grad_norm": 0.67041015625,
|
| 239 |
+
"learning_rate": 0.00012012012012012013,
|
| 240 |
+
"loss": 0.2681,
|
| 241 |
+
"step": 400
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.366863905325444,
|
| 245 |
+
"eval_loss": 0.7119271159172058,
|
| 246 |
+
"eval_runtime": 5.4471,
|
| 247 |
+
"eval_samples_per_second": 16.339,
|
| 248 |
+
"eval_steps_per_second": 2.203,
|
| 249 |
+
"step": 400
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.5147928994082838,
|
| 253 |
+
"grad_norm": 0.841796875,
|
| 254 |
+
"learning_rate": 0.00011511511511511512,
|
| 255 |
+
"loss": 0.2792,
|
| 256 |
+
"step": 425
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.5147928994082838,
|
| 260 |
+
"eval_loss": 0.7170438170433044,
|
| 261 |
+
"eval_runtime": 5.4708,
|
| 262 |
+
"eval_samples_per_second": 16.268,
|
| 263 |
+
"eval_steps_per_second": 2.193,
|
| 264 |
+
"step": 425
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 2.662721893491124,
|
| 268 |
+
"grad_norm": 0.75341796875,
|
| 269 |
+
"learning_rate": 0.00011011011011011012,
|
| 270 |
+
"loss": 0.2738,
|
| 271 |
+
"step": 450
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 2.662721893491124,
|
| 275 |
+
"eval_loss": 0.7100588083267212,
|
| 276 |
+
"eval_runtime": 5.4533,
|
| 277 |
+
"eval_samples_per_second": 16.32,
|
| 278 |
+
"eval_steps_per_second": 2.201,
|
| 279 |
+
"step": 450
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.8106508875739644,
|
| 283 |
+
"grad_norm": 0.8544921875,
|
| 284 |
+
"learning_rate": 0.00010510510510510511,
|
| 285 |
+
"loss": 0.2787,
|
| 286 |
+
"step": 475
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 2.8106508875739644,
|
| 290 |
+
"eval_loss": 0.7120051383972168,
|
| 291 |
+
"eval_runtime": 5.4731,
|
| 292 |
+
"eval_samples_per_second": 16.261,
|
| 293 |
+
"eval_steps_per_second": 2.193,
|
| 294 |
+
"step": 475
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 2.9585798816568047,
|
| 298 |
+
"grad_norm": 0.6982421875,
|
| 299 |
+
"learning_rate": 0.00010010010010010012,
|
| 300 |
+
"loss": 0.2794,
|
| 301 |
+
"step": 500
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 2.9585798816568047,
|
| 305 |
+
"eval_loss": 0.7013982534408569,
|
| 306 |
+
"eval_runtime": 5.4689,
|
| 307 |
+
"eval_samples_per_second": 16.274,
|
| 308 |
+
"eval_steps_per_second": 2.194,
|
| 309 |
+
"step": 500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 3.106508875739645,
|
| 313 |
+
"grad_norm": 0.61669921875,
|
| 314 |
+
"learning_rate": 9.50950950950951e-05,
|
| 315 |
+
"loss": 0.1912,
|
| 316 |
+
"step": 525
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 3.106508875739645,
|
| 320 |
+
"eval_loss": 0.8123311400413513,
|
| 321 |
+
"eval_runtime": 5.4645,
|
| 322 |
+
"eval_samples_per_second": 16.287,
|
| 323 |
+
"eval_steps_per_second": 2.196,
|
| 324 |
+
"step": 525
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 3.2544378698224854,
|
| 328 |
+
"grad_norm": 0.72412109375,
|
| 329 |
+
"learning_rate": 9.009009009009009e-05,
|
| 330 |
+
"loss": 0.1592,
|
| 331 |
+
"step": 550
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 3.2544378698224854,
|
| 335 |
+
"eval_loss": 0.8475367426872253,
|
| 336 |
+
"eval_runtime": 5.4516,
|
| 337 |
+
"eval_samples_per_second": 16.325,
|
| 338 |
+
"eval_steps_per_second": 2.201,
|
| 339 |
+
"step": 550
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 3.4023668639053253,
|
| 343 |
+
"grad_norm": 0.75341796875,
|
| 344 |
+
"learning_rate": 8.50850850850851e-05,
|
| 345 |
+
"loss": 0.1614,
|
| 346 |
+
"step": 575
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 3.4023668639053253,
|
| 350 |
+
"eval_loss": 0.8443621397018433,
|
| 351 |
+
"eval_runtime": 5.4523,
|
| 352 |
+
"eval_samples_per_second": 16.323,
|
| 353 |
+
"eval_steps_per_second": 2.201,
|
| 354 |
+
"step": 575
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 3.5502958579881656,
|
| 358 |
+
"grad_norm": 0.7294921875,
|
| 359 |
+
"learning_rate": 8.008008008008009e-05,
|
| 360 |
+
"loss": 0.1628,
|
| 361 |
+
"step": 600
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 3.5502958579881656,
|
| 365 |
+
"eval_loss": 0.8429920077323914,
|
| 366 |
+
"eval_runtime": 5.4655,
|
| 367 |
+
"eval_samples_per_second": 16.284,
|
| 368 |
+
"eval_steps_per_second": 2.196,
|
| 369 |
+
"step": 600
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 3.698224852071006,
|
| 373 |
+
"grad_norm": 0.90869140625,
|
| 374 |
+
"learning_rate": 7.507507507507507e-05,
|
| 375 |
+
"loss": 0.1599,
|
| 376 |
+
"step": 625
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 3.698224852071006,
|
| 380 |
+
"eval_loss": 0.8433002829551697,
|
| 381 |
+
"eval_runtime": 5.4723,
|
| 382 |
+
"eval_samples_per_second": 16.264,
|
| 383 |
+
"eval_steps_per_second": 2.193,
|
| 384 |
+
"step": 625
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 3.8461538461538463,
|
| 388 |
+
"grad_norm": 0.9658203125,
|
| 389 |
+
"learning_rate": 7.007007007007007e-05,
|
| 390 |
+
"loss": 0.1667,
|
| 391 |
+
"step": 650
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 3.8461538461538463,
|
| 395 |
+
"eval_loss": 0.8468146324157715,
|
| 396 |
+
"eval_runtime": 5.4573,
|
| 397 |
+
"eval_samples_per_second": 16.308,
|
| 398 |
+
"eval_steps_per_second": 2.199,
|
| 399 |
+
"step": 650
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 3.994082840236686,
|
| 403 |
+
"grad_norm": 0.806640625,
|
| 404 |
+
"learning_rate": 6.506506506506507e-05,
|
| 405 |
+
"loss": 0.1643,
|
| 406 |
+
"step": 675
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 3.994082840236686,
|
| 410 |
+
"eval_loss": 0.858656644821167,
|
| 411 |
+
"eval_runtime": 5.4565,
|
| 412 |
+
"eval_samples_per_second": 16.311,
|
| 413 |
+
"eval_steps_per_second": 2.199,
|
| 414 |
+
"step": 675
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 4.1420118343195265,
|
| 418 |
+
"grad_norm": 0.55029296875,
|
| 419 |
+
"learning_rate": 6.0060060060060066e-05,
|
| 420 |
+
"loss": 0.0958,
|
| 421 |
+
"step": 700
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 4.1420118343195265,
|
| 425 |
+
"eval_loss": 0.9950948357582092,
|
| 426 |
+
"eval_runtime": 5.5007,
|
| 427 |
+
"eval_samples_per_second": 16.18,
|
| 428 |
+
"eval_steps_per_second": 2.182,
|
| 429 |
+
"step": 700
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 4.289940828402367,
|
| 433 |
+
"grad_norm": 0.6220703125,
|
| 434 |
+
"learning_rate": 5.505505505505506e-05,
|
| 435 |
+
"loss": 0.0872,
|
| 436 |
+
"step": 725
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 4.289940828402367,
|
| 440 |
+
"eval_loss": 1.022282600402832,
|
| 441 |
+
"eval_runtime": 5.4552,
|
| 442 |
+
"eval_samples_per_second": 16.315,
|
| 443 |
+
"eval_steps_per_second": 2.2,
|
| 444 |
+
"step": 725
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 4.437869822485207,
|
| 448 |
+
"grad_norm": 0.74267578125,
|
| 449 |
+
"learning_rate": 5.005005005005006e-05,
|
| 450 |
+
"loss": 0.0892,
|
| 451 |
+
"step": 750
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 4.437869822485207,
|
| 455 |
+
"eval_loss": 1.0246338844299316,
|
| 456 |
+
"eval_runtime": 5.4786,
|
| 457 |
+
"eval_samples_per_second": 16.245,
|
| 458 |
+
"eval_steps_per_second": 2.19,
|
| 459 |
+
"step": 750
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 4.585798816568047,
|
| 463 |
+
"grad_norm": 0.69970703125,
|
| 464 |
+
"learning_rate": 4.5045045045045046e-05,
|
| 465 |
+
"loss": 0.0892,
|
| 466 |
+
"step": 775
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 4.585798816568047,
|
| 470 |
+
"eval_loss": 1.037864327430725,
|
| 471 |
+
"eval_runtime": 5.5314,
|
| 472 |
+
"eval_samples_per_second": 16.09,
|
| 473 |
+
"eval_steps_per_second": 2.169,
|
| 474 |
+
"step": 775
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.733727810650888,
|
| 478 |
+
"grad_norm": 0.5244140625,
|
| 479 |
+
"learning_rate": 4.0040040040040046e-05,
|
| 480 |
+
"loss": 0.086,
|
| 481 |
+
"step": 800
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.733727810650888,
|
| 485 |
+
"eval_loss": 1.0276521444320679,
|
| 486 |
+
"eval_runtime": 5.5056,
|
| 487 |
+
"eval_samples_per_second": 16.165,
|
| 488 |
+
"eval_steps_per_second": 2.18,
|
| 489 |
+
"step": 800
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.881656804733728,
|
| 493 |
+
"grad_norm": 0.7646484375,
|
| 494 |
+
"learning_rate": 3.503503503503503e-05,
|
| 495 |
+
"loss": 0.0871,
|
| 496 |
+
"step": 825
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.881656804733728,
|
| 500 |
+
"eval_loss": 1.058767318725586,
|
| 501 |
+
"eval_runtime": 5.4583,
|
| 502 |
+
"eval_samples_per_second": 16.305,
|
| 503 |
+
"eval_steps_per_second": 2.198,
|
| 504 |
+
"step": 825
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 5.029585798816568,
|
| 508 |
+
"grad_norm": 0.45458984375,
|
| 509 |
+
"learning_rate": 3.0030030030030033e-05,
|
| 510 |
+
"loss": 0.0808,
|
| 511 |
+
"step": 850
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 5.029585798816568,
|
| 515 |
+
"eval_loss": 1.081122636795044,
|
| 516 |
+
"eval_runtime": 5.4567,
|
| 517 |
+
"eval_samples_per_second": 16.31,
|
| 518 |
+
"eval_steps_per_second": 2.199,
|
| 519 |
+
"step": 850
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 5.177514792899408,
|
| 523 |
+
"grad_norm": 0.5234375,
|
| 524 |
+
"learning_rate": 2.502502502502503e-05,
|
| 525 |
+
"loss": 0.0512,
|
| 526 |
+
"step": 875
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 5.177514792899408,
|
| 530 |
+
"eval_loss": 1.168296217918396,
|
| 531 |
+
"eval_runtime": 5.4565,
|
| 532 |
+
"eval_samples_per_second": 16.311,
|
| 533 |
+
"eval_steps_per_second": 2.199,
|
| 534 |
+
"step": 875
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 5.325443786982248,
|
| 538 |
+
"grad_norm": 0.43701171875,
|
| 539 |
+
"learning_rate": 2.0020020020020023e-05,
|
| 540 |
+
"loss": 0.0512,
|
| 541 |
+
"step": 900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 5.325443786982248,
|
| 545 |
+
"eval_loss": 1.161507487297058,
|
| 546 |
+
"eval_runtime": 5.5102,
|
| 547 |
+
"eval_samples_per_second": 16.152,
|
| 548 |
+
"eval_steps_per_second": 2.178,
|
| 549 |
+
"step": 900
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 5.4733727810650885,
|
| 553 |
+
"grad_norm": 0.56298828125,
|
| 554 |
+
"learning_rate": 1.5015015015015016e-05,
|
| 555 |
+
"loss": 0.051,
|
| 556 |
+
"step": 925
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 5.4733727810650885,
|
| 560 |
+
"eval_loss": 1.1937472820281982,
|
| 561 |
+
"eval_runtime": 5.5151,
|
| 562 |
+
"eval_samples_per_second": 16.138,
|
| 563 |
+
"eval_steps_per_second": 2.176,
|
| 564 |
+
"step": 925
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 5.621301775147929,
|
| 568 |
+
"grad_norm": 0.56396484375,
|
| 569 |
+
"learning_rate": 1.0010010010010011e-05,
|
| 570 |
+
"loss": 0.0493,
|
| 571 |
+
"step": 950
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 5.621301775147929,
|
| 575 |
+
"eval_loss": 1.184690237045288,
|
| 576 |
+
"eval_runtime": 5.4656,
|
| 577 |
+
"eval_samples_per_second": 16.284,
|
| 578 |
+
"eval_steps_per_second": 2.196,
|
| 579 |
+
"step": 950
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 5.769230769230769,
|
| 583 |
+
"grad_norm": 0.43017578125,
|
| 584 |
+
"learning_rate": 5.005005005005006e-06,
|
| 585 |
+
"loss": 0.0496,
|
| 586 |
+
"step": 975
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 5.769230769230769,
|
| 590 |
+
"eval_loss": 1.1814072132110596,
|
| 591 |
+
"eval_runtime": 5.4875,
|
| 592 |
+
"eval_samples_per_second": 16.219,
|
| 593 |
+
"eval_steps_per_second": 2.187,
|
| 594 |
+
"step": 975
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 5.9171597633136095,
|
| 598 |
+
"grad_norm": 0.56103515625,
|
| 599 |
+
"learning_rate": 0.0,
|
| 600 |
+
"loss": 0.0469,
|
| 601 |
+
"step": 1000
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 5.9171597633136095,
|
| 605 |
+
"eval_loss": 1.1828829050064087,
|
| 606 |
+
"eval_runtime": 5.4573,
|
| 607 |
+
"eval_samples_per_second": 16.308,
|
| 608 |
+
"eval_steps_per_second": 2.199,
|
| 609 |
+
"step": 1000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 5.9171597633136095,
|
| 613 |
+
"step": 1000,
|
| 614 |
+
"total_flos": 1.75885655212032e+17,
|
| 615 |
+
"train_loss": 0.27826149678230283,
|
| 616 |
+
"train_runtime": 1217.2703,
|
| 617 |
+
"train_samples_per_second": 3.286,
|
| 618 |
+
"train_steps_per_second": 0.822
|
| 619 |
+
}
|
| 620 |
+
],
|
| 621 |
+
"logging_steps": 25,
|
| 622 |
+
"max_steps": 1000,
|
| 623 |
+
"num_input_tokens_seen": 0,
|
| 624 |
+
"num_train_epochs": 6,
|
| 625 |
+
"save_steps": 25,
|
| 626 |
+
"total_flos": 1.75885655212032e+17,
|
| 627 |
+
"train_batch_size": 1,
|
| 628 |
+
"trial_name": null,
|
| 629 |
+
"trial_params": null
|
| 630 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a7c5ec06d43296f0b9931d5829d30fca0fd6033a27c2147cd9e3b913ec6ee18
|
| 3 |
+
size 4920
|