Hugging Face's logo

bartowski
/
starchat2-15b-v0.1-exl2

Text Generation
alignment-handbook
Generated from Trainer
Model card Files Community
bartowski commited on
Commit
5dea819
•
1 Parent(s): ffb5c47

Quant for 4.25

Browse files
Files changed (19) hide show
  1. .gitattributes +1 -0
  2. README.md +125 -43
  3. added_tokens.json +4 -0
  4. all_results.json +21 -0
  5. config.json +31 -0
  6. eval_results.json +16 -0
  7. generation_config.json +6 -0
  8. merges.txt +0 -0
  9. model.safetensors.index.json +651 -0
  10. model_logo.png +3 -0
  11. original_repo_url.txt +1 -0
  12. output-00001-of-00002.safetensors +3 -0
  13. output-00002-of-00002.safetensors +3 -0
  14. special_tokens_map.json +34 -0
  15. tokenizer.json +0 -0
  16. tokenizer_config.json +338 -0
  17. train_results.json +8 -0
  18. trainer_state.json +1931 -0
  19. vocab.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ model_logo.png filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -9,69 +9,151 @@ datasets:
9
  model-index:
10
  - name: starchat2-15b-v0.1
11
  results: []
12
- quantized_by: bartowski
13
- pipeline_tag: text-generation
14
  ---
15
 
16
- ## Exllama v2 Quantizations of starchat2-15b-v0.1
17
 
18
- Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.15">turboderp's ExLlamaV2 v0.0.15</a> for quantization.
19
 
20
- ## The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)
21
 
22
- Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
23
 
24
- Conversion was done using the default calibration dataset.
25
 
26
- Default arguments used except when the bits per weight is above 6.0, at that point the lm_head layer is quantized at 8 bits per weight instead of the default 6.
27
 
28
- Original model: https://huggingface.co/HuggingFaceH4/starchat2-15b-v0.1
 
 
 
29
 
 
30
 
31
- <a href="https://huggingface.co/bartowski/starchat2-15b-v0.1-exl2/tree/8_0">8.0 bits per weight</a>
32
 
33
- <a href="https://huggingface.co/bartowski/starchat2-15b-v0.1-exl2/tree/6_5">6.5 bits per weight</a>
 
34
 
35
- <a href="https://huggingface.co/bartowski/starchat2-15b-v0.1-exl2/tree/5_0">5.0 bits per weight</a>
36
 
37
- <a href="https://huggingface.co/bartowski/starchat2-15b-v0.1-exl2/tree/4_25">4.25 bits per weight</a>
38
 
39
- <a href="https://huggingface.co/bartowski/starchat2-15b-v0.1-exl2/tree/3_5">3.5 bits per weight</a>
 
 
 
 
40
 
41
 
42
- ## Download instructions
43
 
44
- With git:
45
 
46
- ```shell
47
- git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/starchat2-15b-v0.1-exl2
48
- ```
49
-
50
- With huggingface hub (credit to TheBloke for instructions):
51
-
52
- ```shell
53
- pip3 install huggingface-hub
54
- ```
55
 
56
- To download the `main` (only useful if you only care about measurement.json) branch to a folder called `starchat2-15b-v0.1-exl2`:
 
 
57
 
58
- ```shell
59
- mkdir starchat2-15b-v0.1-exl2
60
- huggingface-cli download bartowski/starchat2-15b-v0.1-exl2 --local-dir starchat2-15b-v0.1-exl2 --local-dir-use-symlinks False
61
- ```
62
-
63
- To download from a different branch, add the `--revision` parameter:
64
 
65
- Linux:
66
-
67
- ```shell
68
- mkdir starchat2-15b-v0.1-exl2-6_5
69
- huggingface-cli download bartowski/starchat2-15b-v0.1-exl2 --revision 6_5 --local-dir starchat2-15b-v0.1-exl2-6_5 --local-dir-use-symlinks False
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
70
  ```
71
 
72
- Windows (which apparently doesn't like _ in folders sometimes?):
73
-
74
- ```shell
75
- mkdir starchat2-15b-v0.1-exl2-6.5
76
- huggingface-cli download bartowski/starchat2-15b-v0.1-exl2 --revision 6_5 --local-dir starchat2-15b-v0.1-exl2-6.5 --local-dir-use-symlinks False
77
- ```
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9
  model-index:
10
  - name: starchat2-15b-v0.1
11
  results: []
 
 
12
  ---
13
 
14
+ <img src="https://huggingface.co/HuggingFaceH4/starchat2-15b-v0.1/resolve/main/model_logo.png" alt="StarChat2 15B Logo" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
15
 
16
+ # Model Card for StarChat2 15B
17
 
18
+ StarChat is a series of language models that are trained to act as helpful coding assistants. StarChat2 is the latest model in the series, and is a fine-tuned version of [StarCoder2](https://huggingface.co/bigcode/starcoder2-15b) that was trained with SFT and DPO on a mix of synthetic datasets.
19
 
20
+ ## Model Details
21
 
22
+ ### Model Description
23
 
24
+ <!-- Provide a longer summary of what this model is. -->
25
 
26
+ - **Model type:** A 16B parameter GPT-like model fine-tuned on a mix of publicly available, synthetic datasets.
27
+ - **Language(s) (NLP):** Primarily English and 80+ programming languages.
28
+ - **License:** BigCode Open RAIL-M v1
29
+ - **Finetuned from model:** [bigcode/starcoder2-15b](https://huggingface.co/bigcode/starcoder2-15b)
30
 
31
+ ### Model Sources
32
 
33
+ <!-- Provide the basic links for the model. -->
34
 
35
+ - **Repository:** https://github.com/huggingface/alignment-handbook
36
+ - **Demo:** https://huggingface.co/spaces/HuggingFaceH4/starchat2-playground
37
 
38
+ ## Performance
39
 
40
+ StarChat2 15B was trained to balance chat and programming capabilities. It achieves strong performance on chat benchmarks like [MT Bench](https://huggingface.co/spaces/lmsys/mt-bench) and [IFEval](https://arxiv.org/abs/2311.07911), as well as the canonical HumanEval benchmark for Python code completion. The scores reported below were obtained using the [LightEval](https://github.com/huggingface/lighteval) evaluation suite (commit `988959cb905df4baa050f82b4d499d46e8b537f2`) and each prompt has been formatted with the model's corresponding chat template to simulate real-world usage. This is why some scores may differ from those reported in technical reports or on the Open LLM Leaderboard.
41
 
42
+ | Model | MT Bench | IFEval | HumanEval |
43
+ |-------------------------------------------------------------------------------------------------|---------:|-------:|----------:|
44
+ | [starchat2-15b-v0.1](https://huggingface.co/HuggingFaceH4/starchat2-15b-v0.1) | 7.66 | 35.12 | 71.34 |
45
+ | [deepseek-coder-6.7b-instruct](https://huggingface.co/deepseek-ai/deepseek-coder-6.7b-instruct) | 4.17 | 14.23 | 80.48 |
46
+ | [CodeLlama-13b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-13b-Instruct-hf) | 6.80 | 43.44 | 50.60 |
47
 
48
 
49
+ ## Intended uses & limitations
50
 
51
+ The model was fine-tuned on a blend of chat, code, math, and reasoning datasets. As a result, the model can be used for chat and you can check out our [demo](https://huggingface.co/spaces/HuggingFaceH4/starchat2-playground) to test its coding capabilities.
52
 
53
+ Here's how you can run the model using the `pipeline()` function from 🤗 Transformers:
 
 
 
 
 
 
 
 
54
 
55
+ ```python
56
+ # pip install 'transformers @ git+https://github.com/huggingface/transformers.git@831bc25d8fdb85768402f772cf65cc3d7872b211'
57
+ # pip install accelerate
58
 
59
+ import torch
60
+ from transformers import pipeline
 
 
 
 
61
 
62
+ pipe = pipeline(
63
+ "text-generation",
64
+ model="HuggingFaceH4/starchat2-15b-v0.1",
65
+ device_map="auto",
66
+ torch_dtype=torch.bfloat16,
67
+ )
68
+ messages = [
69
+ {
70
+ "role": "system",
71
+ "content": "You are StarChat2, an expert programming assistant",
72
+ },
73
+ {"role": "user", "content": "Write a simple website in HTML. When a user clicks the button, it shows a random Chuck Norris joke."},
74
+ ]
75
+ outputs = pipe(
76
+ messages,
77
+ max_new_tokens=512,
78
+ do_sample=True,
79
+ temperature=0.7,
80
+ top_k=50,
81
+ top_p=0.95,
82
+ stop_sequence="<|im_end|>",
83
+ )
84
+ print(outputs[0]["generated_text"][-1]["content"])
85
  ```
86
 
87
+ ## Bias, Risks, and Limitations
88
+
89
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
90
+
91
+ StarChat2 15B has not been aligned to human preferences with techniques like RLHF or deployed with in-the-loop filtering of responses like ChatGPT, so the model can produce problematic outputs (especially when prompted to do so).
92
+ Models trained primarily on code data will also have a more skewed demographic bias commensurate with the demographics of the GitHub community, for more on this see the [StarCoder2 dataset](https://huggingface.co/datasets/bigcode/the-stack-v2)
93
+
94
+ Since the base model was pretrained on a large corpus of code, it may produce code snippets that are syntactically valid but semantically incorrect.
95
+ For example, it may produce code that does not compile or that produces incorrect results.
96
+ It may also produce code that is vulnerable to security exploits.
97
+ We have observed the model also has a tendency to produce false URLs which should be carefully inspected before clicking.
98
+
99
+ StarChat2 15B was fine-tuned from the base model [StarCoder2](https://huggingface.co/bigcode/starcoder2-15b), please refer to its model card's [Limitations Section](https://huggingface.co/bigcode/starcoder2-15b#limitations) for relevant information.
100
+ In particular, the model was evaluated on some categories of gender biases, propensity for toxicity, and risk of suggesting code completions with known security flaws; these evaluations are reported in its [technical report](https://huggingface.co/papers/2402.19173).
101
+
102
+
103
+ ## Training details
104
+
105
+ This model is a fine-tuned version of [starchat2-15b-sft-v0.1](https://huggingface.co/HuggingFaceH4/starchat2-15b-sft-v0.1) on the HuggingFaceH4/ultrafeedback_binarized and the HuggingFaceH4/orca_dpo_pairs datasets. Check out the recipe in the [Alignment Handbook](https://github.com/huggingface/alignment-handbook) for more details.
106
+
107
+ It achieves the following results on the evaluation set:
108
+ - Loss: 0.4347
109
+ - Rewards/chosen: -0.9461
110
+ - Rewards/rejected: -2.7745
111
+ - Rewards/accuracies: 0.7658
112
+ - Rewards/margins: 1.8284
113
+ - Logps/rejected: -322.1934
114
+ - Logps/chosen: -316.1898
115
+ - Logits/rejected: -2.3817
116
+ - Logits/chosen: -2.3005
117
+
118
+ ## Training procedure
119
+
120
+ ### Training hyperparameters
121
+
122
+ The following hyperparameters were used during training:
123
+ - learning_rate: 5e-07
124
+ - train_batch_size: 2
125
+ - eval_batch_size: 4
126
+ - seed: 42
127
+ - distributed_type: multi-GPU
128
+ - num_devices: 8
129
+ - gradient_accumulation_steps: 8
130
+ - total_train_batch_size: 128
131
+ - total_eval_batch_size: 32
132
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
133
+ - lr_scheduler_type: cosine
134
+ - lr_scheduler_warmup_ratio: 0.1
135
+ - num_epochs: 2
136
+
137
+ ### Training results
138
+
139
+ | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
140
+ |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
141
+ | 0.717 | 0.17 | 100 | 0.6006 | -0.0924 | -0.2899 | 0.6329 | 0.1975 | -272.5022 | -299.1165 | -2.5313 | -2.4191 |
142
+ | 0.6273 | 0.35 | 200 | 0.5160 | -0.3994 | -0.9461 | 0.6930 | 0.5467 | -285.6261 | -305.2568 | -2.5281 | -2.4278 |
143
+ | 0.5538 | 0.52 | 300 | 0.4781 | -0.6589 | -1.5892 | 0.7247 | 0.9302 | -298.4870 | -310.4470 | -2.4996 | -2.4110 |
144
+ | 0.5056 | 0.7 | 400 | 0.4594 | -0.8283 | -2.1332 | 0.7437 | 1.3050 | -309.3687 | -313.8344 | -2.4472 | -2.3644 |
145
+ | 0.4983 | 0.87 | 500 | 0.4512 | -0.7758 | -2.2806 | 0.7468 | 1.5049 | -312.3167 | -312.7843 | -2.4223 | -2.3404 |
146
+ | 0.4662 | 1.04 | 600 | 0.4431 | -0.7839 | -2.4016 | 0.7658 | 1.6177 | -314.7355 | -312.9465 | -2.4049 | -2.3215 |
147
+ | 0.4411 | 1.22 | 700 | 0.4415 | -1.0090 | -2.7582 | 0.7690 | 1.7492 | -321.8679 | -317.4481 | -2.3840 | -2.3016 |
148
+ | 0.471 | 1.39 | 800 | 0.4368 | -0.9617 | -2.7445 | 0.7690 | 1.7828 | -321.5930 | -316.5019 | -2.3809 | -2.2991 |
149
+ | 0.4485 | 1.57 | 900 | 0.4351 | -0.9490 | -2.7594 | 0.7722 | 1.8103 | -321.8916 | -316.2497 | -2.3815 | -2.3004 |
150
+ | 0.4411 | 1.74 | 1000 | 0.4348 | -0.9293 | -2.7469 | 0.7658 | 1.8176 | -321.6409 | -315.8547 | -2.3823 | -2.3011 |
151
+ | 0.4499 | 1.92 | 1100 | 0.4348 | -0.9482 | -2.7767 | 0.7658 | 1.8285 | -322.2369 | -316.2320 | -2.3828 | -2.3012 |
152
+
153
+
154
+ ### Framework versions
155
+
156
+ - Transformers 4.39.0.dev0
157
+ - Pytorch 2.1.2+cu121
158
+ - Datasets 2.16.1
159
+ - Tokenizers 0.15.1
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<|im_end|>": 49153,
3
+ "<|im_start|>": 49152
4
+ }
all_results.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "eval_logits/chosen": -2.3005340099334717,
4
+ "eval_logits/rejected": -2.3817031383514404,
5
+ "eval_logps/chosen": -316.1898498535156,
6
+ "eval_logps/rejected": -322.1933898925781,
7
+ "eval_loss": 0.4346597194671631,
8
+ "eval_rewards/accuracies": 0.7658227682113647,
9
+ "eval_rewards/chosen": -0.9460535049438477,
10
+ "eval_rewards/margins": 1.8284220695495605,
11
+ "eval_rewards/rejected": -2.7744758129119873,
12
+ "eval_runtime": 117.6177,
13
+ "eval_samples": 2500,
14
+ "eval_samples_per_second": 21.255,
15
+ "eval_steps_per_second": 0.672,
16
+ "train_loss": 0.5184940074794384,
17
+ "train_runtime": 19743.6623,
18
+ "train_samples": 73494,
19
+ "train_samples_per_second": 7.445,
20
+ "train_steps_per_second": 0.058
21
+ }
config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "HuggingFaceH4/starcoder2-15b-ift",
3
+ "architectures": [
4
+ "Starcoder2ForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.1,
7
+ "bos_token_id": 0,
8
+ "embedding_dropout": 0.1,
9
+ "eos_token_id": 0,
10
+ "hidden_act": "gelu_pytorch_tanh",
11
+ "hidden_size": 6144,
12
+ "initializer_range": 0.01275,
13
+ "intermediate_size": 24576,
14
+ "max_position_embeddings": 16384,
15
+ "mlp_type": "default",
16
+ "model_type": "starcoder2",
17
+ "norm_epsilon": 1e-05,
18
+ "norm_type": "layer_norm",
19
+ "num_attention_heads": 48,
20
+ "num_hidden_layers": 40,
21
+ "num_key_value_heads": 4,
22
+ "residual_dropout": 0.1,
23
+ "rope_theta": 100000,
24
+ "sliding_window": 4096,
25
+ "tie_word_embeddings": false,
26
+ "torch_dtype": "bfloat16",
27
+ "transformers_version": "4.39.0.dev0",
28
+ "use_bias": true,
29
+ "use_cache": true,
30
+ "vocab_size": 49154
31
+ }
eval_results.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "eval_logits/chosen": -2.3005340099334717,
4
+ "eval_logits/rejected": -2.3817031383514404,
5
+ "eval_logps/chosen": -316.1898498535156,
6
+ "eval_logps/rejected": -322.1933898925781,
7
+ "eval_loss": 0.4346597194671631,
8
+ "eval_rewards/accuracies": 0.7658227682113647,
9
+ "eval_rewards/chosen": -0.9460535049438477,
10
+ "eval_rewards/margins": 1.8284220695495605,
11
+ "eval_rewards/rejected": -2.7744758129119873,
12
+ "eval_runtime": 117.6177,
13
+ "eval_samples": 2500,
14
+ "eval_samples_per_second": 21.255,
15
+ "eval_steps_per_second": 0.672
16
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.39.0.dev0"
6
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model.safetensors.index.json ADDED
@@ -0,0 +1,651 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 31915827200
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "model-00007-of-00007.safetensors",
7
+ "model.embed_tokens.weight": "model-00001-of-00007.safetensors",
8
+ "model.layers.0.input_layernorm.bias": "model-00001-of-00007.safetensors",
9
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00007.safetensors",
10
+ "model.layers.0.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
11
+ "model.layers.0.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
12
+ "model.layers.0.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
13
+ "model.layers.0.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
14
+ "model.layers.0.post_attention_layernorm.bias": "model-00001-of-00007.safetensors",
15
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00007.safetensors",
16
+ "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
17
+ "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
18
+ "model.layers.0.self_attn.o_proj.bias": "model-00001-of-00007.safetensors",
19
+ "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
20
+ "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
21
+ "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
22
+ "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
23
+ "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
24
+ "model.layers.1.input_layernorm.bias": "model-00001-of-00007.safetensors",
25
+ "model.layers.1.input_layernorm.weight": "model-00001-of-00007.safetensors",
26
+ "model.layers.1.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
27
+ "model.layers.1.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
28
+ "model.layers.1.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
29
+ "model.layers.1.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
30
+ "model.layers.1.post_attention_layernorm.bias": "model-00001-of-00007.safetensors",
31
+ "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00007.safetensors",
32
+ "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
33
+ "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
34
+ "model.layers.1.self_attn.o_proj.bias": "model-00001-of-00007.safetensors",
35
+ "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
36
+ "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
37
+ "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
38
+ "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
39
+ "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
40
+ "model.layers.10.input_layernorm.bias": "model-00002-of-00007.safetensors",
41
+ "model.layers.10.input_layernorm.weight": "model-00002-of-00007.safetensors",
42
+ "model.layers.10.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
43
+ "model.layers.10.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
44
+ "model.layers.10.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
45
+ "model.layers.10.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
46
+ "model.layers.10.post_attention_layernorm.bias": "model-00002-of-00007.safetensors",
47
+ "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
48
+ "model.layers.10.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
49
+ "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
50
+ "model.layers.10.self_attn.o_proj.bias": "model-00002-of-00007.safetensors",
51
+ "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
52
+ "model.layers.10.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
53
+ "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
54
+ "model.layers.10.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
55
+ "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
56
+ "model.layers.11.input_layernorm.bias": "model-00002-of-00007.safetensors",
57
+ "model.layers.11.input_layernorm.weight": "model-00002-of-00007.safetensors",
58
+ "model.layers.11.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
59
+ "model.layers.11.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
60
+ "model.layers.11.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
61
+ "model.layers.11.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
62
+ "model.layers.11.post_attention_layernorm.bias": "model-00002-of-00007.safetensors",
63
+ "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
64
+ "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
65
+ "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
66
+ "model.layers.11.self_attn.o_proj.bias": "model-00002-of-00007.safetensors",
67
+ "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
68
+ "model.layers.11.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
69
+ "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
70
+ "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
71
+ "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
72
+ "model.layers.12.input_layernorm.bias": "model-00003-of-00007.safetensors",
73
+ "model.layers.12.input_layernorm.weight": "model-00003-of-00007.safetensors",
74
+ "model.layers.12.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
75
+ "model.layers.12.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
76
+ "model.layers.12.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
77
+ "model.layers.12.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
78
+ "model.layers.12.post_attention_layernorm.bias": "model-00003-of-00007.safetensors",
79
+ "model.layers.12.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
80
+ "model.layers.12.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
81
+ "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
82
+ "model.layers.12.self_attn.o_proj.bias": "model-00003-of-00007.safetensors",
83
+ "model.layers.12.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
84
+ "model.layers.12.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
85
+ "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
86
+ "model.layers.12.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
87
+ "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
88
+ "model.layers.13.input_layernorm.bias": "model-00003-of-00007.safetensors",
89
+ "model.layers.13.input_layernorm.weight": "model-00003-of-00007.safetensors",
90
+ "model.layers.13.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
91
+ "model.layers.13.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
92
+ "model.layers.13.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
93
+ "model.layers.13.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
94
+ "model.layers.13.post_attention_layernorm.bias": "model-00003-of-00007.safetensors",
95
+ "model.layers.13.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
96
+ "model.layers.13.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
97
+ "model.layers.13.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
98
+ "model.layers.13.self_attn.o_proj.bias": "model-00003-of-00007.safetensors",
99
+ "model.layers.13.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
100
+ "model.layers.13.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
101
+ "model.layers.13.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
102
+ "model.layers.13.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
103
+ "model.layers.13.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
104
+ "model.layers.14.input_layernorm.bias": "model-00003-of-00007.safetensors",
105
+ "model.layers.14.input_layernorm.weight": "model-00003-of-00007.safetensors",
106
+ "model.layers.14.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
107
+ "model.layers.14.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
108
+ "model.layers.14.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
109
+ "model.layers.14.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
110
+ "model.layers.14.post_attention_layernorm.bias": "model-00003-of-00007.safetensors",
111
+ "model.layers.14.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
112
+ "model.layers.14.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
113
+ "model.layers.14.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
114
+ "model.layers.14.self_attn.o_proj.bias": "model-00003-of-00007.safetensors",
115
+ "model.layers.14.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
116
+ "model.layers.14.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
117
+ "model.layers.14.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
118
+ "model.layers.14.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
119
+ "model.layers.14.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
120
+ "model.layers.15.input_layernorm.bias": "model-00003-of-00007.safetensors",
121
+ "model.layers.15.input_layernorm.weight": "model-00003-of-00007.safetensors",
122
+ "model.layers.15.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
123
+ "model.layers.15.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
124
+ "model.layers.15.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
125
+ "model.layers.15.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
126
+ "model.layers.15.post_attention_layernorm.bias": "model-00003-of-00007.safetensors",
127
+ "model.layers.15.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
128
+ "model.layers.15.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
129
+ "model.layers.15.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
130
+ "model.layers.15.self_attn.o_proj.bias": "model-00003-of-00007.safetensors",
131
+ "model.layers.15.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
132
+ "model.layers.15.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
133
+ "model.layers.15.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
134
+ "model.layers.15.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
135
+ "model.layers.15.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
136
+ "model.layers.16.input_layernorm.bias": "model-00003-of-00007.safetensors",
137
+ "model.layers.16.input_layernorm.weight": "model-00003-of-00007.safetensors",
138
+ "model.layers.16.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
139
+ "model.layers.16.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
140
+ "model.layers.16.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
141
+ "model.layers.16.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
142
+ "model.layers.16.post_attention_layernorm.bias": "model-00003-of-00007.safetensors",
143
+ "model.layers.16.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
144
+ "model.layers.16.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
145
+ "model.layers.16.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
146
+ "model.layers.16.self_attn.o_proj.bias": "model-00003-of-00007.safetensors",
147
+ "model.layers.16.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
148
+ "model.layers.16.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
149
+ "model.layers.16.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
150
+ "model.layers.16.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
151
+ "model.layers.16.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
152
+ "model.layers.17.input_layernorm.bias": "model-00003-of-00007.safetensors",
153
+ "model.layers.17.input_layernorm.weight": "model-00003-of-00007.safetensors",
154
+ "model.layers.17.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
155
+ "model.layers.17.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
156
+ "model.layers.17.mlp.c_proj.bias": "model-00003-of-00007.safetensors",
157
+ "model.layers.17.mlp.c_proj.weight": "model-00003-of-00007.safetensors",
158
+ "model.layers.17.post_attention_layernorm.bias": "model-00003-of-00007.safetensors",
159
+ "model.layers.17.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
160
+ "model.layers.17.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
161
+ "model.layers.17.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
162
+ "model.layers.17.self_attn.o_proj.bias": "model-00003-of-00007.safetensors",
163
+ "model.layers.17.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
164
+ "model.layers.17.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
165
+ "model.layers.17.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
166
+ "model.layers.17.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
167
+ "model.layers.17.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
168
+ "model.layers.18.input_layernorm.bias": "model-00004-of-00007.safetensors",
169
+ "model.layers.18.input_layernorm.weight": "model-00004-of-00007.safetensors",
170
+ "model.layers.18.mlp.c_fc.bias": "model-00003-of-00007.safetensors",
171
+ "model.layers.18.mlp.c_fc.weight": "model-00003-of-00007.safetensors",
172
+ "model.layers.18.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
173
+ "model.layers.18.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
174
+ "model.layers.18.post_attention_layernorm.bias": "model-00004-of-00007.safetensors",
175
+ "model.layers.18.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
176
+ "model.layers.18.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
177
+ "model.layers.18.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
178
+ "model.layers.18.self_attn.o_proj.bias": "model-00003-of-00007.safetensors",
179
+ "model.layers.18.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
180
+ "model.layers.18.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
181
+ "model.layers.18.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
182
+ "model.layers.18.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
183
+ "model.layers.18.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
184
+ "model.layers.19.input_layernorm.bias": "model-00004-of-00007.safetensors",
185
+ "model.layers.19.input_layernorm.weight": "model-00004-of-00007.safetensors",
186
+ "model.layers.19.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
187
+ "model.layers.19.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
188
+ "model.layers.19.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
189
+ "model.layers.19.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
190
+ "model.layers.19.post_attention_layernorm.bias": "model-00004-of-00007.safetensors",
191
+ "model.layers.19.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
192
+ "model.layers.19.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
193
+ "model.layers.19.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
194
+ "model.layers.19.self_attn.o_proj.bias": "model-00004-of-00007.safetensors",
195
+ "model.layers.19.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
196
+ "model.layers.19.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
197
+ "model.layers.19.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
198
+ "model.layers.19.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
199
+ "model.layers.19.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
200
+ "model.layers.2.input_layernorm.bias": "model-00001-of-00007.safetensors",
201
+ "model.layers.2.input_layernorm.weight": "model-00001-of-00007.safetensors",
202
+ "model.layers.2.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
203
+ "model.layers.2.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
204
+ "model.layers.2.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
205
+ "model.layers.2.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
206
+ "model.layers.2.post_attention_layernorm.bias": "model-00001-of-00007.safetensors",
207
+ "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00007.safetensors",
208
+ "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
209
+ "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
210
+ "model.layers.2.self_attn.o_proj.bias": "model-00001-of-00007.safetensors",
211
+ "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
212
+ "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
213
+ "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
214
+ "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
215
+ "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
216
+ "model.layers.20.input_layernorm.bias": "model-00004-of-00007.safetensors",
217
+ "model.layers.20.input_layernorm.weight": "model-00004-of-00007.safetensors",
218
+ "model.layers.20.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
219
+ "model.layers.20.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
220
+ "model.layers.20.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
221
+ "model.layers.20.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
222
+ "model.layers.20.post_attention_layernorm.bias": "model-00004-of-00007.safetensors",
223
+ "model.layers.20.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
224
+ "model.layers.20.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
225
+ "model.layers.20.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
226
+ "model.layers.20.self_attn.o_proj.bias": "model-00004-of-00007.safetensors",
227
+ "model.layers.20.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
228
+ "model.layers.20.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
229
+ "model.layers.20.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
230
+ "model.layers.20.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
231
+ "model.layers.20.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
232
+ "model.layers.21.input_layernorm.bias": "model-00004-of-00007.safetensors",
233
+ "model.layers.21.input_layernorm.weight": "model-00004-of-00007.safetensors",
234
+ "model.layers.21.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
235
+ "model.layers.21.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
236
+ "model.layers.21.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
237
+ "model.layers.21.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
238
+ "model.layers.21.post_attention_layernorm.bias": "model-00004-of-00007.safetensors",
239
+ "model.layers.21.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
240
+ "model.layers.21.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
241
+ "model.layers.21.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
242
+ "model.layers.21.self_attn.o_proj.bias": "model-00004-of-00007.safetensors",
243
+ "model.layers.21.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
244
+ "model.layers.21.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
245
+ "model.layers.21.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
246
+ "model.layers.21.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
247
+ "model.layers.21.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
248
+ "model.layers.22.input_layernorm.bias": "model-00004-of-00007.safetensors",
249
+ "model.layers.22.input_layernorm.weight": "model-00004-of-00007.safetensors",
250
+ "model.layers.22.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
251
+ "model.layers.22.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
252
+ "model.layers.22.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
253
+ "model.layers.22.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
254
+ "model.layers.22.post_attention_layernorm.bias": "model-00004-of-00007.safetensors",
255
+ "model.layers.22.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
256
+ "model.layers.22.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
257
+ "model.layers.22.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
258
+ "model.layers.22.self_attn.o_proj.bias": "model-00004-of-00007.safetensors",
259
+ "model.layers.22.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
260
+ "model.layers.22.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
261
+ "model.layers.22.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
262
+ "model.layers.22.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
263
+ "model.layers.22.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
264
+ "model.layers.23.input_layernorm.bias": "model-00004-of-00007.safetensors",
265
+ "model.layers.23.input_layernorm.weight": "model-00004-of-00007.safetensors",
266
+ "model.layers.23.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
267
+ "model.layers.23.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
268
+ "model.layers.23.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
269
+ "model.layers.23.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
270
+ "model.layers.23.post_attention_layernorm.bias": "model-00004-of-00007.safetensors",
271
+ "model.layers.23.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
272
+ "model.layers.23.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
273
+ "model.layers.23.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
274
+ "model.layers.23.self_attn.o_proj.bias": "model-00004-of-00007.safetensors",
275
+ "model.layers.23.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
276
+ "model.layers.23.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
277
+ "model.layers.23.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
278
+ "model.layers.23.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
279
+ "model.layers.23.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
280
+ "model.layers.24.input_layernorm.bias": "model-00004-of-00007.safetensors",
281
+ "model.layers.24.input_layernorm.weight": "model-00004-of-00007.safetensors",
282
+ "model.layers.24.mlp.c_fc.bias": "model-00004-of-00007.safetensors",
283
+ "model.layers.24.mlp.c_fc.weight": "model-00004-of-00007.safetensors",
284
+ "model.layers.24.mlp.c_proj.bias": "model-00004-of-00007.safetensors",
285
+ "model.layers.24.mlp.c_proj.weight": "model-00004-of-00007.safetensors",
286
+ "model.layers.24.post_attention_layernorm.bias": "model-00004-of-00007.safetensors",
287
+ "model.layers.24.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
288
+ "model.layers.24.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
289
+ "model.layers.24.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
290
+ "model.layers.24.self_attn.o_proj.bias": "model-00004-of-00007.safetensors",
291
+ "model.layers.24.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
292
+ "model.layers.24.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
293
+ "model.layers.24.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
294
+ "model.layers.24.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
295
+ "model.layers.24.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
296
+ "model.layers.25.input_layernorm.bias": "model-00005-of-00007.safetensors",
297
+ "model.layers.25.input_layernorm.weight": "model-00005-of-00007.safetensors",
298
+ "model.layers.25.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
299
+ "model.layers.25.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
300
+ "model.layers.25.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
301
+ "model.layers.25.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
302
+ "model.layers.25.post_attention_layernorm.bias": "model-00005-of-00007.safetensors",
303
+ "model.layers.25.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
304
+ "model.layers.25.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
305
+ "model.layers.25.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
306
+ "model.layers.25.self_attn.o_proj.bias": "model-00005-of-00007.safetensors",
307
+ "model.layers.25.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
308
+ "model.layers.25.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
309
+ "model.layers.25.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
310
+ "model.layers.25.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
311
+ "model.layers.25.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
312
+ "model.layers.26.input_layernorm.bias": "model-00005-of-00007.safetensors",
313
+ "model.layers.26.input_layernorm.weight": "model-00005-of-00007.safetensors",
314
+ "model.layers.26.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
315
+ "model.layers.26.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
316
+ "model.layers.26.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
317
+ "model.layers.26.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
318
+ "model.layers.26.post_attention_layernorm.bias": "model-00005-of-00007.safetensors",
319
+ "model.layers.26.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
320
+ "model.layers.26.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
321
+ "model.layers.26.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
322
+ "model.layers.26.self_attn.o_proj.bias": "model-00005-of-00007.safetensors",
323
+ "model.layers.26.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
324
+ "model.layers.26.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
325
+ "model.layers.26.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
326
+ "model.layers.26.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
327
+ "model.layers.26.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
328
+ "model.layers.27.input_layernorm.bias": "model-00005-of-00007.safetensors",
329
+ "model.layers.27.input_layernorm.weight": "model-00005-of-00007.safetensors",
330
+ "model.layers.27.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
331
+ "model.layers.27.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
332
+ "model.layers.27.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
333
+ "model.layers.27.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
334
+ "model.layers.27.post_attention_layernorm.bias": "model-00005-of-00007.safetensors",
335
+ "model.layers.27.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
336
+ "model.layers.27.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
337
+ "model.layers.27.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
338
+ "model.layers.27.self_attn.o_proj.bias": "model-00005-of-00007.safetensors",
339
+ "model.layers.27.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
340
+ "model.layers.27.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
341
+ "model.layers.27.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
342
+ "model.layers.27.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
343
+ "model.layers.27.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
344
+ "model.layers.28.input_layernorm.bias": "model-00005-of-00007.safetensors",
345
+ "model.layers.28.input_layernorm.weight": "model-00005-of-00007.safetensors",
346
+ "model.layers.28.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
347
+ "model.layers.28.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
348
+ "model.layers.28.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
349
+ "model.layers.28.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
350
+ "model.layers.28.post_attention_layernorm.bias": "model-00005-of-00007.safetensors",
351
+ "model.layers.28.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
352
+ "model.layers.28.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
353
+ "model.layers.28.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
354
+ "model.layers.28.self_attn.o_proj.bias": "model-00005-of-00007.safetensors",
355
+ "model.layers.28.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
356
+ "model.layers.28.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
357
+ "model.layers.28.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
358
+ "model.layers.28.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
359
+ "model.layers.28.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
360
+ "model.layers.29.input_layernorm.bias": "model-00005-of-00007.safetensors",
361
+ "model.layers.29.input_layernorm.weight": "model-00005-of-00007.safetensors",
362
+ "model.layers.29.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
363
+ "model.layers.29.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
364
+ "model.layers.29.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
365
+ "model.layers.29.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
366
+ "model.layers.29.post_attention_layernorm.bias": "model-00005-of-00007.safetensors",
367
+ "model.layers.29.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
368
+ "model.layers.29.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
369
+ "model.layers.29.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
370
+ "model.layers.29.self_attn.o_proj.bias": "model-00005-of-00007.safetensors",
371
+ "model.layers.29.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
372
+ "model.layers.29.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
373
+ "model.layers.29.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
374
+ "model.layers.29.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
375
+ "model.layers.29.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
376
+ "model.layers.3.input_layernorm.bias": "model-00001-of-00007.safetensors",
377
+ "model.layers.3.input_layernorm.weight": "model-00001-of-00007.safetensors",
378
+ "model.layers.3.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
379
+ "model.layers.3.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
380
+ "model.layers.3.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
381
+ "model.layers.3.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
382
+ "model.layers.3.post_attention_layernorm.bias": "model-00001-of-00007.safetensors",
383
+ "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00007.safetensors",
384
+ "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
385
+ "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
386
+ "model.layers.3.self_attn.o_proj.bias": "model-00001-of-00007.safetensors",
387
+ "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
388
+ "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
389
+ "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
390
+ "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
391
+ "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
392
+ "model.layers.30.input_layernorm.bias": "model-00005-of-00007.safetensors",
393
+ "model.layers.30.input_layernorm.weight": "model-00005-of-00007.safetensors",
394
+ "model.layers.30.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
395
+ "model.layers.30.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
396
+ "model.layers.30.mlp.c_proj.bias": "model-00005-of-00007.safetensors",
397
+ "model.layers.30.mlp.c_proj.weight": "model-00005-of-00007.safetensors",
398
+ "model.layers.30.post_attention_layernorm.bias": "model-00005-of-00007.safetensors",
399
+ "model.layers.30.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
400
+ "model.layers.30.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
401
+ "model.layers.30.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
402
+ "model.layers.30.self_attn.o_proj.bias": "model-00005-of-00007.safetensors",
403
+ "model.layers.30.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
404
+ "model.layers.30.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
405
+ "model.layers.30.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
406
+ "model.layers.30.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
407
+ "model.layers.30.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
408
+ "model.layers.31.input_layernorm.bias": "model-00006-of-00007.safetensors",
409
+ "model.layers.31.input_layernorm.weight": "model-00006-of-00007.safetensors",
410
+ "model.layers.31.mlp.c_fc.bias": "model-00005-of-00007.safetensors",
411
+ "model.layers.31.mlp.c_fc.weight": "model-00005-of-00007.safetensors",
412
+ "model.layers.31.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
413
+ "model.layers.31.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
414
+ "model.layers.31.post_attention_layernorm.bias": "model-00006-of-00007.safetensors",
415
+ "model.layers.31.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
416
+ "model.layers.31.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
417
+ "model.layers.31.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
418
+ "model.layers.31.self_attn.o_proj.bias": "model-00005-of-00007.safetensors",
419
+ "model.layers.31.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
420
+ "model.layers.31.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
421
+ "model.layers.31.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
422
+ "model.layers.31.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
423
+ "model.layers.31.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
424
+ "model.layers.32.input_layernorm.bias": "model-00006-of-00007.safetensors",
425
+ "model.layers.32.input_layernorm.weight": "model-00006-of-00007.safetensors",
426
+ "model.layers.32.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
427
+ "model.layers.32.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
428
+ "model.layers.32.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
429
+ "model.layers.32.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
430
+ "model.layers.32.post_attention_layernorm.bias": "model-00006-of-00007.safetensors",
431
+ "model.layers.32.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
432
+ "model.layers.32.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
433
+ "model.layers.32.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
434
+ "model.layers.32.self_attn.o_proj.bias": "model-00006-of-00007.safetensors",
435
+ "model.layers.32.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
436
+ "model.layers.32.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
437
+ "model.layers.32.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
438
+ "model.layers.32.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
439
+ "model.layers.32.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
440
+ "model.layers.33.input_layernorm.bias": "model-00006-of-00007.safetensors",
441
+ "model.layers.33.input_layernorm.weight": "model-00006-of-00007.safetensors",
442
+ "model.layers.33.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
443
+ "model.layers.33.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
444
+ "model.layers.33.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
445
+ "model.layers.33.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
446
+ "model.layers.33.post_attention_layernorm.bias": "model-00006-of-00007.safetensors",
447
+ "model.layers.33.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
448
+ "model.layers.33.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
449
+ "model.layers.33.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
450
+ "model.layers.33.self_attn.o_proj.bias": "model-00006-of-00007.safetensors",
451
+ "model.layers.33.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
452
+ "model.layers.33.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
453
+ "model.layers.33.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
454
+ "model.layers.33.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
455
+ "model.layers.33.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
456
+ "model.layers.34.input_layernorm.bias": "model-00006-of-00007.safetensors",
457
+ "model.layers.34.input_layernorm.weight": "model-00006-of-00007.safetensors",
458
+ "model.layers.34.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
459
+ "model.layers.34.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
460
+ "model.layers.34.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
461
+ "model.layers.34.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
462
+ "model.layers.34.post_attention_layernorm.bias": "model-00006-of-00007.safetensors",
463
+ "model.layers.34.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
464
+ "model.layers.34.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
465
+ "model.layers.34.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
466
+ "model.layers.34.self_attn.o_proj.bias": "model-00006-of-00007.safetensors",
467
+ "model.layers.34.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
468
+ "model.layers.34.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
469
+ "model.layers.34.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
470
+ "model.layers.34.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
471
+ "model.layers.34.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
472
+ "model.layers.35.input_layernorm.bias": "model-00006-of-00007.safetensors",
473
+ "model.layers.35.input_layernorm.weight": "model-00006-of-00007.safetensors",
474
+ "model.layers.35.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
475
+ "model.layers.35.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
476
+ "model.layers.35.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
477
+ "model.layers.35.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
478
+ "model.layers.35.post_attention_layernorm.bias": "model-00006-of-00007.safetensors",
479
+ "model.layers.35.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
480
+ "model.layers.35.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
481
+ "model.layers.35.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
482
+ "model.layers.35.self_attn.o_proj.bias": "model-00006-of-00007.safetensors",
483
+ "model.layers.35.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
484
+ "model.layers.35.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
485
+ "model.layers.35.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
486
+ "model.layers.35.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
487
+ "model.layers.35.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
488
+ "model.layers.36.input_layernorm.bias": "model-00006-of-00007.safetensors",
489
+ "model.layers.36.input_layernorm.weight": "model-00006-of-00007.safetensors",
490
+ "model.layers.36.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
491
+ "model.layers.36.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
492
+ "model.layers.36.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
493
+ "model.layers.36.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
494
+ "model.layers.36.post_attention_layernorm.bias": "model-00006-of-00007.safetensors",
495
+ "model.layers.36.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
496
+ "model.layers.36.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
497
+ "model.layers.36.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
498
+ "model.layers.36.self_attn.o_proj.bias": "model-00006-of-00007.safetensors",
499
+ "model.layers.36.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
500
+ "model.layers.36.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
501
+ "model.layers.36.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
502
+ "model.layers.36.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
503
+ "model.layers.36.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
504
+ "model.layers.37.input_layernorm.bias": "model-00006-of-00007.safetensors",
505
+ "model.layers.37.input_layernorm.weight": "model-00006-of-00007.safetensors",
506
+ "model.layers.37.mlp.c_fc.bias": "model-00006-of-00007.safetensors",
507
+ "model.layers.37.mlp.c_fc.weight": "model-00006-of-00007.safetensors",
508
+ "model.layers.37.mlp.c_proj.bias": "model-00006-of-00007.safetensors",
509
+ "model.layers.37.mlp.c_proj.weight": "model-00006-of-00007.safetensors",
510
+ "model.layers.37.post_attention_layernorm.bias": "model-00006-of-00007.safetensors",
511
+ "model.layers.37.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
512
+ "model.layers.37.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
513
+ "model.layers.37.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
514
+ "model.layers.37.self_attn.o_proj.bias": "model-00006-of-00007.safetensors",
515
+ "model.layers.37.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
516
+ "model.layers.37.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
517
+ "model.layers.37.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
518
+ "model.layers.37.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
519
+ "model.layers.37.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
520
+ "model.layers.38.input_layernorm.bias": "model-00007-of-00007.safetensors",
521
+ "model.layers.38.input_layernorm.weight": "model-00007-of-00007.safetensors",
522
+ "model.layers.38.mlp.c_fc.bias": "model-00007-of-00007.safetensors",
523
+ "model.layers.38.mlp.c_fc.weight": "model-00007-of-00007.safetensors",
524
+ "model.layers.38.mlp.c_proj.bias": "model-00007-of-00007.safetensors",
525
+ "model.layers.38.mlp.c_proj.weight": "model-00007-of-00007.safetensors",
526
+ "model.layers.38.post_attention_layernorm.bias": "model-00007-of-00007.safetensors",
527
+ "model.layers.38.post_attention_layernorm.weight": "model-00007-of-00007.safetensors",
528
+ "model.layers.38.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
529
+ "model.layers.38.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
530
+ "model.layers.38.self_attn.o_proj.bias": "model-00007-of-00007.safetensors",
531
+ "model.layers.38.self_attn.o_proj.weight": "model-00007-of-00007.safetensors",
532
+ "model.layers.38.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
533
+ "model.layers.38.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
534
+ "model.layers.38.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
535
+ "model.layers.38.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
536
+ "model.layers.39.input_layernorm.bias": "model-00007-of-00007.safetensors",
537
+ "model.layers.39.input_layernorm.weight": "model-00007-of-00007.safetensors",
538
+ "model.layers.39.mlp.c_fc.bias": "model-00007-of-00007.safetensors",
539
+ "model.layers.39.mlp.c_fc.weight": "model-00007-of-00007.safetensors",
540
+ "model.layers.39.mlp.c_proj.bias": "model-00007-of-00007.safetensors",
541
+ "model.layers.39.mlp.c_proj.weight": "model-00007-of-00007.safetensors",
542
+ "model.layers.39.post_attention_layernorm.bias": "model-00007-of-00007.safetensors",
543
+ "model.layers.39.post_attention_layernorm.weight": "model-00007-of-00007.safetensors",
544
+ "model.layers.39.self_attn.k_proj.bias": "model-00007-of-00007.safetensors",
545
+ "model.layers.39.self_attn.k_proj.weight": "model-00007-of-00007.safetensors",
546
+ "model.layers.39.self_attn.o_proj.bias": "model-00007-of-00007.safetensors",
547
+ "model.layers.39.self_attn.o_proj.weight": "model-00007-of-00007.safetensors",
548
+ "model.layers.39.self_attn.q_proj.bias": "model-00007-of-00007.safetensors",
549
+ "model.layers.39.self_attn.q_proj.weight": "model-00007-of-00007.safetensors",
550
+ "model.layers.39.self_attn.v_proj.bias": "model-00007-of-00007.safetensors",
551
+ "model.layers.39.self_attn.v_proj.weight": "model-00007-of-00007.safetensors",
552
+ "model.layers.4.input_layernorm.bias": "model-00001-of-00007.safetensors",
553
+ "model.layers.4.input_layernorm.weight": "model-00001-of-00007.safetensors",
554
+ "model.layers.4.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
555
+ "model.layers.4.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
556
+ "model.layers.4.mlp.c_proj.bias": "model-00001-of-00007.safetensors",
557
+ "model.layers.4.mlp.c_proj.weight": "model-00001-of-00007.safetensors",
558
+ "model.layers.4.post_attention_layernorm.bias": "model-00001-of-00007.safetensors",
559
+ "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00007.safetensors",
560
+ "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
561
+ "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
562
+ "model.layers.4.self_attn.o_proj.bias": "model-00001-of-00007.safetensors",
563
+ "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
564
+ "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
565
+ "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
566
+ "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
567
+ "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
568
+ "model.layers.5.input_layernorm.bias": "model-00002-of-00007.safetensors",
569
+ "model.layers.5.input_layernorm.weight": "model-00002-of-00007.safetensors",
570
+ "model.layers.5.mlp.c_fc.bias": "model-00001-of-00007.safetensors",
571
+ "model.layers.5.mlp.c_fc.weight": "model-00001-of-00007.safetensors",
572
+ "model.layers.5.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
573
+ "model.layers.5.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
574
+ "model.layers.5.post_attention_layernorm.bias": "model-00002-of-00007.safetensors",
575
+ "model.layers.5.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
576
+ "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
577
+ "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
578
+ "model.layers.5.self_attn.o_proj.bias": "model-00001-of-00007.safetensors",
579
+ "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
580
+ "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
581
+ "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
582
+ "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
583
+ "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
584
+ "model.layers.6.input_layernorm.bias": "model-00002-of-00007.safetensors",
585
+ "model.layers.6.input_layernorm.weight": "model-00002-of-00007.safetensors",
586
+ "model.layers.6.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
587
+ "model.layers.6.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
588
+ "model.layers.6.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
589
+ "model.layers.6.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
590
+ "model.layers.6.post_attention_layernorm.bias": "model-00002-of-00007.safetensors",
591
+ "model.layers.6.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
592
+ "model.layers.6.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
593
+ "model.layers.6.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
594
+ "model.layers.6.self_attn.o_proj.bias": "model-00002-of-00007.safetensors",
595
+ "model.layers.6.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
596
+ "model.layers.6.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
597
+ "model.layers.6.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
598
+ "model.layers.6.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
599
+ "model.layers.6.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
600
+ "model.layers.7.input_layernorm.bias": "model-00002-of-00007.safetensors",
601
+ "model.layers.7.input_layernorm.weight": "model-00002-of-00007.safetensors",
602
+ "model.layers.7.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
603
+ "model.layers.7.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
604
+ "model.layers.7.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
605
+ "model.layers.7.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
606
+ "model.layers.7.post_attention_layernorm.bias": "model-00002-of-00007.safetensors",
607
+ "model.layers.7.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
608
+ "model.layers.7.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
609
+ "model.layers.7.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
610
+ "model.layers.7.self_attn.o_proj.bias": "model-00002-of-00007.safetensors",
611
+ "model.layers.7.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
612
+ "model.layers.7.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
613
+ "model.layers.7.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
614
+ "model.layers.7.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
615
+ "model.layers.7.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
616
+ "model.layers.8.input_layernorm.bias": "model-00002-of-00007.safetensors",
617
+ "model.layers.8.input_layernorm.weight": "model-00002-of-00007.safetensors",
618
+ "model.layers.8.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
619
+ "model.layers.8.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
620
+ "model.layers.8.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
621
+ "model.layers.8.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
622
+ "model.layers.8.post_attention_layernorm.bias": "model-00002-of-00007.safetensors",
623
+ "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
624
+ "model.layers.8.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
625
+ "model.layers.8.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
626
+ "model.layers.8.self_attn.o_proj.bias": "model-00002-of-00007.safetensors",
627
+ "model.layers.8.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
628
+ "model.layers.8.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
629
+ "model.layers.8.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
630
+ "model.layers.8.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
631
+ "model.layers.8.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
632
+ "model.layers.9.input_layernorm.bias": "model-00002-of-00007.safetensors",
633
+ "model.layers.9.input_layernorm.weight": "model-00002-of-00007.safetensors",
634
+ "model.layers.9.mlp.c_fc.bias": "model-00002-of-00007.safetensors",
635
+ "model.layers.9.mlp.c_fc.weight": "model-00002-of-00007.safetensors",
636
+ "model.layers.9.mlp.c_proj.bias": "model-00002-of-00007.safetensors",
637
+ "model.layers.9.mlp.c_proj.weight": "model-00002-of-00007.safetensors",
638
+ "model.layers.9.post_attention_layernorm.bias": "model-00002-of-00007.safetensors",
639
+ "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
640
+ "model.layers.9.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
641
+ "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
642
+ "model.layers.9.self_attn.o_proj.bias": "model-00002-of-00007.safetensors",
643
+ "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
644
+ "model.layers.9.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
645
+ "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
646
+ "model.layers.9.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
647
+ "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
648
+ "model.norm.bias": "model-00007-of-00007.safetensors",
649
+ "model.norm.weight": "model-00007-of-00007.safetensors"
650
+ }
651
+ }
model_logo.png ADDED

Git LFS Details

  • SHA256: 85603fd9cdff6d10e73147bfc3ed2c07cd7e8054a44d295bc2bebd6c51837948
  • Pointer size: 132 Bytes
  • Size of remote file: 3.14 MB
original_repo_url.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ https://huggingface.co/HuggingFaceH4/starchat2-15b-v0.1
output-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97d275effe92e950587a4e2ee6af84f009669b6e22a2158e46c3d1a058a438ae
3
+ size 8581606848
output-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42071cd8a35c744b2d88f6b456dc0ff35f6d49e8ae76380340299d8ad1a901d9
3
+ size 419591000
special_tokens_map.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>"
5
+ ],
6
+ "bos_token": {
7
+ "content": "<|im_start|>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false
12
+ },
13
+ "eos_token": {
14
+ "content": "<|im_end|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false
19
+ },
20
+ "pad_token": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false
26
+ },
27
+ "unk_token": {
28
+ "content": "<|endoftext|>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false
33
+ }
34
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,338 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<fim_prefix>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "<fim_middle>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "<fim_suffix>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "4": {
37
+ "content": "<fim_pad>",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "5": {
45
+ "content": "<repo_name>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": true
51
+ },
52
+ "6": {
53
+ "content": "<file_sep>",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": true
59
+ },
60
+ "7": {
61
+ "content": "<issue_start>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false,
66
+ "special": true
67
+ },
68
+ "8": {
69
+ "content": "<issue_comment>",
70
+ "lstrip": false,
71
+ "normalized": false,
72
+ "rstrip": false,
73
+ "single_word": false,
74
+ "special": true
75
+ },
76
+ "9": {
77
+ "content": "<issue_closed>",
78
+ "lstrip": false,
79
+ "normalized": false,
80
+ "rstrip": false,
81
+ "single_word": false,
82
+ "special": true
83
+ },
84
+ "10": {
85
+ "content": "<jupyter_start>",
86
+ "lstrip": false,
87
+ "normalized": false,
88
+ "rstrip": false,
89
+ "single_word": false,
90
+ "special": true
91
+ },
92
+ "11": {
93
+ "content": "<jupyter_text>",
94
+ "lstrip": false,
95
+ "normalized": false,
96
+ "rstrip": false,
97
+ "single_word": false,
98
+ "special": true
99
+ },
100
+ "12": {
101
+ "content": "<jupyter_code>",
102
+ "lstrip": false,
103
+ "normalized": false,
104
+ "rstrip": false,
105
+ "single_word": false,
106
+ "special": true
107
+ },
108
+ "13": {
109
+ "content": "<jupyter_output>",
110
+ "lstrip": false,
111
+ "normalized": false,
112
+ "rstrip": false,
113
+ "single_word": false,
114
+ "special": true
115
+ },
116
+ "14": {
117
+ "content": "<jupyter_script>",
118
+ "lstrip": false,
119
+ "normalized": false,
120
+ "rstrip": false,
121
+ "single_word": false,
122
+ "special": true
123
+ },
124
+ "15": {
125
+ "content": "<empty_output>",
126
+ "lstrip": false,
127
+ "normalized": false,
128
+ "rstrip": false,
129
+ "single_word": false,
130
+ "special": true
131
+ },
132
+ "16": {
133
+ "content": "<code_to_intermediate>",
134
+ "lstrip": false,
135
+ "normalized": false,
136
+ "rstrip": false,
137
+ "single_word": false,
138
+ "special": true
139
+ },
140
+ "17": {
141
+ "content": "<intermediate_to_code>",
142
+ "lstrip": false,
143
+ "normalized": false,
144
+ "rstrip": false,
145
+ "single_word": false,
146
+ "special": true
147
+ },
148
+ "18": {
149
+ "content": "<pr>",
150
+ "lstrip": false,
151
+ "normalized": false,
152
+ "rstrip": false,
153
+ "single_word": false,
154
+ "special": true
155
+ },
156
+ "19": {
157
+ "content": "<pr_status>",
158
+ "lstrip": false,
159
+ "normalized": false,
160
+ "rstrip": false,
161
+ "single_word": false,
162
+ "special": true
163
+ },
164
+ "20": {
165
+ "content": "<pr_is_merged>",
166
+ "lstrip": false,
167
+ "normalized": false,
168
+ "rstrip": false,
169
+ "single_word": false,
170
+ "special": true
171
+ },
172
+ "21": {
173
+ "content": "<pr_base>",
174
+ "lstrip": false,
175
+ "normalized": false,
176
+ "rstrip": false,
177
+ "single_word": false,
178
+ "special": true
179
+ },
180
+ "22": {
181
+ "content": "<pr_file>",
182
+ "lstrip": false,
183
+ "normalized": false,
184
+ "rstrip": false,
185
+ "single_word": false,
186
+ "special": true
187
+ },
188
+ "23": {
189
+ "content": "<pr_base_code>",
190
+ "lstrip": false,
191
+ "normalized": false,
192
+ "rstrip": false,
193
+ "single_word": false,
194
+ "special": true
195
+ },
196
+ "24": {
197
+ "content": "<pr_diff>",
198
+ "lstrip": false,
199
+ "normalized": false,
200
+ "rstrip": false,
201
+ "single_word": false,
202
+ "special": true
203
+ },
204
+ "25": {
205
+ "content": "<pr_diff_hunk>",
206
+ "lstrip": false,
207
+ "normalized": false,
208
+ "rstrip": false,
209
+ "single_word": false,
210
+ "special": true
211
+ },
212
+ "26": {
213
+ "content": "<pr_comment>",
214
+ "lstrip": false,
215
+ "normalized": false,
216
+ "rstrip": false,
217
+ "single_word": false,
218
+ "special": true
219
+ },
220
+ "27": {
221
+ "content": "<pr_event_id>",
222
+ "lstrip": false,
223
+ "normalized": false,
224
+ "rstrip": false,
225
+ "single_word": false,
226
+ "special": true
227
+ },
228
+ "28": {
229
+ "content": "<pr_review>",
230
+ "lstrip": false,
231
+ "normalized": false,
232
+ "rstrip": false,
233
+ "single_word": false,
234
+ "special": true
235
+ },
236
+ "29": {
237
+ "content": "<pr_review_state>",
238
+ "lstrip": false,
239
+ "normalized": false,
240
+ "rstrip": false,
241
+ "single_word": false,
242
+ "special": true
243
+ },
244
+ "30": {
245
+ "content": "<pr_review_comment>",
246
+ "lstrip": false,
247
+ "normalized": false,
248
+ "rstrip": false,
249
+ "single_word": false,
250
+ "special": true
251
+ },
252
+ "31": {
253
+ "content": "<pr_in_reply_to_review_id>",
254
+ "lstrip": false,
255
+ "normalized": false,
256
+ "rstrip": false,
257
+ "single_word": false,
258
+ "special": true
259
+ },
260
+ "32": {
261
+ "content": "<pr_in_reply_to_comment_id>",
262
+ "lstrip": false,
263
+ "normalized": false,
264
+ "rstrip": false,
265
+ "single_word": false,
266
+ "special": true
267
+ },
268
+ "33": {
269
+ "content": "<pr_diff_hunk_comment_line>",
270
+ "lstrip": false,
271
+ "normalized": false,
272
+ "rstrip": false,
273
+ "single_word": false,
274
+ "special": true
275
+ },
276
+ "34": {
277
+ "content": "<NAME>",
278
+ "lstrip": false,
279
+ "normalized": false,
280
+ "rstrip": false,
281
+ "single_word": false,
282
+ "special": true
283
+ },
284
+ "35": {
285
+ "content": "<EMAIL>",
286
+ "lstrip": false,
287
+ "normalized": false,
288
+ "rstrip": false,
289
+ "single_word": false,
290
+ "special": true
291
+ },
292
+ "36": {
293
+ "content": "<KEY>",
294
+ "lstrip": false,
295
+ "normalized": false,
296
+ "rstrip": false,
297
+ "single_word": false,
298
+ "special": true
299
+ },
300
+ "37": {
301
+ "content": "<PASSWORD>",
302
+ "lstrip": false,
303
+ "normalized": false,
304
+ "rstrip": false,
305
+ "single_word": false,
306
+ "special": true
307
+ },
308
+ "49152": {
309
+ "content": "<|im_start|>",
310
+ "lstrip": false,
311
+ "normalized": false,
312
+ "rstrip": false,
313
+ "single_word": false,
314
+ "special": true
315
+ },
316
+ "49153": {
317
+ "content": "<|im_end|>",
318
+ "lstrip": false,
319
+ "normalized": false,
320
+ "rstrip": false,
321
+ "single_word": false,
322
+ "special": true
323
+ }
324
+ },
325
+ "additional_special_tokens": [
326
+ "<|im_start|>",
327
+ "<|im_end|>"
328
+ ],
329
+ "bos_token": "<|im_start|>",
330
+ "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 %}",
331
+ "clean_up_tokenization_spaces": true,
332
+ "eos_token": "<|im_end|>",
333
+ "model_max_length": 2048,
334
+ "pad_token": "<|im_end|>",
335
+ "tokenizer_class": "GPT2Tokenizer",
336
+ "unk_token": "<|endoftext|>",
337
+ "vocab_size": 49152
338
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "train_loss": 0.5184940074794384,
4
+ "train_runtime": 19743.6623,
5
+ "train_samples": 73494,
6
+ "train_samples_per_second": 7.445,
7
+ "train_steps_per_second": 0.058
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1931 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.999129299085764,
5
+ "eval_steps": 100,
6
+ "global_step": 1148,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "grad_norm": 8.541543291462938,
14
+ "learning_rate": 4.347826086956522e-09,
15
+ "logits/chosen": -2.3435652256011963,
16
+ "logits/rejected": -2.551011562347412,
17
+ "logps/chosen": -314.6017150878906,
18
+ "logps/rejected": -206.73678588867188,
19
+ "loss": 0.7439,
20
+ "rewards/accuracies": 0.4375,
21
+ "rewards/chosen": -0.2740848958492279,
22
+ "rewards/margins": -0.1701626181602478,
23
+ "rewards/rejected": -0.1039222925901413,
24
+ "step": 1
25
+ },
26
+ {
27
+ "epoch": 0.02,
28
+ "grad_norm": 8.82551006058537,
29
+ "learning_rate": 4.347826086956521e-08,
30
+ "logits/chosen": -2.4006946086883545,
31
+ "logits/rejected": -2.4231932163238525,
32
+ "logps/chosen": -289.26739501953125,
33
+ "logps/rejected": -272.8511962890625,
34
+ "loss": 0.7793,
35
+ "rewards/accuracies": 0.3888888955116272,
36
+ "rewards/chosen": -0.30362409353256226,
37
+ "rewards/margins": -0.10056456923484802,
38
+ "rewards/rejected": -0.20305952429771423,
39
+ "step": 10
40
+ },
41
+ {
42
+ "epoch": 0.03,
43
+ "grad_norm": 9.957017780765279,
44
+ "learning_rate": 8.695652173913042e-08,
45
+ "logits/chosen": -2.366244077682495,
46
+ "logits/rejected": -2.436676263809204,
47
+ "logps/chosen": -304.1680908203125,
48
+ "logps/rejected": -283.8580017089844,
49
+ "loss": 0.7817,
50
+ "rewards/accuracies": 0.5249999761581421,
51
+ "rewards/chosen": -0.31308335065841675,
52
+ "rewards/margins": -0.0013480648631229997,
53
+ "rewards/rejected": -0.3117353320121765,
54
+ "step": 20
55
+ },
56
+ {
57
+ "epoch": 0.05,
58
+ "grad_norm": 9.441280424617158,
59
+ "learning_rate": 1.3043478260869563e-07,
60
+ "logits/chosen": -2.4001221656799316,
61
+ "logits/rejected": -2.398263454437256,
62
+ "logps/chosen": -299.45147705078125,
63
+ "logps/rejected": -312.4380187988281,
64
+ "loss": 0.808,
65
+ "rewards/accuracies": 0.4312500059604645,
66
+ "rewards/chosen": -0.40518712997436523,
67
+ "rewards/margins": -0.170148104429245,
68
+ "rewards/rejected": -0.23503902554512024,
69
+ "step": 30
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "grad_norm": 9.0042393863242,
74
+ "learning_rate": 1.7391304347826085e-07,
75
+ "logits/chosen": -2.309999704360962,
76
+ "logits/rejected": -2.3476955890655518,
77
+ "logps/chosen": -297.22235107421875,
78
+ "logps/rejected": -302.00482177734375,
79
+ "loss": 0.7707,
80
+ "rewards/accuracies": 0.53125,
81
+ "rewards/chosen": -0.2780631482601166,
82
+ "rewards/margins": 0.08002600818872452,
83
+ "rewards/rejected": -0.3580891489982605,
84
+ "step": 40
85
+ },
86
+ {
87
+ "epoch": 0.09,
88
+ "grad_norm": 9.03964644769571,
89
+ "learning_rate": 2.1739130434782607e-07,
90
+ "logits/chosen": -2.3683385848999023,
91
+ "logits/rejected": -2.3768067359924316,
92
+ "logps/chosen": -288.1864013671875,
93
+ "logps/rejected": -275.24493408203125,
94
+ "loss": 0.7913,
95
+ "rewards/accuracies": 0.4312500059604645,
96
+ "rewards/chosen": -0.24404355883598328,
97
+ "rewards/margins": -0.0613979808986187,
98
+ "rewards/rejected": -0.18264558911323547,
99
+ "step": 50
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "grad_norm": 9.171931303155361,
104
+ "learning_rate": 2.6086956521739126e-07,
105
+ "logits/chosen": -2.3965792655944824,
106
+ "logits/rejected": -2.4396941661834717,
107
+ "logps/chosen": -313.60186767578125,
108
+ "logps/rejected": -277.5401611328125,
109
+ "loss": 0.7896,
110
+ "rewards/accuracies": 0.42500001192092896,
111
+ "rewards/chosen": -0.3664000332355499,
112
+ "rewards/margins": -0.11313207447528839,
113
+ "rewards/rejected": -0.25326794385910034,
114
+ "step": 60
115
+ },
116
+ {
117
+ "epoch": 0.12,
118
+ "grad_norm": 9.303112541467058,
119
+ "learning_rate": 3.043478260869565e-07,
120
+ "logits/chosen": -2.3145995140075684,
121
+ "logits/rejected": -2.3265597820281982,
122
+ "logps/chosen": -288.7591247558594,
123
+ "logps/rejected": -282.3059997558594,
124
+ "loss": 0.7541,
125
+ "rewards/accuracies": 0.44999998807907104,
126
+ "rewards/chosen": -0.3928884267807007,
127
+ "rewards/margins": -0.10098680108785629,
128
+ "rewards/rejected": -0.2919016480445862,
129
+ "step": 70
130
+ },
131
+ {
132
+ "epoch": 0.14,
133
+ "grad_norm": 8.148953826123538,
134
+ "learning_rate": 3.478260869565217e-07,
135
+ "logits/chosen": -2.4307453632354736,
136
+ "logits/rejected": -2.408132553100586,
137
+ "logps/chosen": -264.6097106933594,
138
+ "logps/rejected": -279.7698669433594,
139
+ "loss": 0.7471,
140
+ "rewards/accuracies": 0.48750001192092896,
141
+ "rewards/chosen": -0.32603269815444946,
142
+ "rewards/margins": 0.050616730004549026,
143
+ "rewards/rejected": -0.3766494393348694,
144
+ "step": 80
145
+ },
146
+ {
147
+ "epoch": 0.16,
148
+ "grad_norm": 8.978142275251097,
149
+ "learning_rate": 3.9130434782608694e-07,
150
+ "logits/chosen": -2.3797852993011475,
151
+ "logits/rejected": -2.370492935180664,
152
+ "logps/chosen": -272.8327331542969,
153
+ "logps/rejected": -269.51495361328125,
154
+ "loss": 0.7172,
155
+ "rewards/accuracies": 0.5625,
156
+ "rewards/chosen": -0.3148571252822876,
157
+ "rewards/margins": 0.12950611114501953,
158
+ "rewards/rejected": -0.44436320662498474,
159
+ "step": 90
160
+ },
161
+ {
162
+ "epoch": 0.17,
163
+ "grad_norm": 8.101534684245069,
164
+ "learning_rate": 4.3478260869565214e-07,
165
+ "logits/chosen": -2.3395884037017822,
166
+ "logits/rejected": -2.3768749237060547,
167
+ "logps/chosen": -302.3770751953125,
168
+ "logps/rejected": -289.23370361328125,
169
+ "loss": 0.717,
170
+ "rewards/accuracies": 0.543749988079071,
171
+ "rewards/chosen": -0.4626343846321106,
172
+ "rewards/margins": 0.1476312279701233,
173
+ "rewards/rejected": -0.6102656126022339,
174
+ "step": 100
175
+ },
176
+ {
177
+ "epoch": 0.17,
178
+ "eval_logits/chosen": -2.4191040992736816,
179
+ "eval_logits/rejected": -2.5313072204589844,
180
+ "eval_logps/chosen": -299.11651611328125,
181
+ "eval_logps/rejected": -272.5022277832031,
182
+ "eval_loss": 0.6005984544754028,
183
+ "eval_rewards/accuracies": 0.6329113841056824,
184
+ "eval_rewards/chosen": -0.09238887578248978,
185
+ "eval_rewards/margins": 0.19752810895442963,
186
+ "eval_rewards/rejected": -0.2899169921875,
187
+ "eval_runtime": 116.6781,
188
+ "eval_samples_per_second": 21.426,
189
+ "eval_steps_per_second": 0.677,
190
+ "step": 100
191
+ },
192
+ {
193
+ "epoch": 0.19,
194
+ "grad_norm": 9.267934110515439,
195
+ "learning_rate": 4.782608695652174e-07,
196
+ "logits/chosen": -2.402156114578247,
197
+ "logits/rejected": -2.372450351715088,
198
+ "logps/chosen": -310.5009460449219,
199
+ "logps/rejected": -306.57513427734375,
200
+ "loss": 0.7206,
201
+ "rewards/accuracies": 0.550000011920929,
202
+ "rewards/chosen": -0.4795568585395813,
203
+ "rewards/margins": 0.2181818038225174,
204
+ "rewards/rejected": -0.697738528251648,
205
+ "step": 110
206
+ },
207
+ {
208
+ "epoch": 0.21,
209
+ "grad_norm": 7.237175384641583,
210
+ "learning_rate": 4.99971097144235e-07,
211
+ "logits/chosen": -2.3216421604156494,
212
+ "logits/rejected": -2.3850231170654297,
213
+ "logps/chosen": -299.10943603515625,
214
+ "logps/rejected": -275.91632080078125,
215
+ "loss": 0.6797,
216
+ "rewards/accuracies": 0.6000000238418579,
217
+ "rewards/chosen": -0.4861680567264557,
218
+ "rewards/margins": 0.2638343870639801,
219
+ "rewards/rejected": -0.7500024437904358,
220
+ "step": 120
221
+ },
222
+ {
223
+ "epoch": 0.23,
224
+ "grad_norm": 8.125054397906975,
225
+ "learning_rate": 4.997399143945734e-07,
226
+ "logits/chosen": -2.319718599319458,
227
+ "logits/rejected": -2.338841438293457,
228
+ "logps/chosen": -293.934326171875,
229
+ "logps/rejected": -285.8225402832031,
230
+ "loss": 0.6779,
231
+ "rewards/accuracies": 0.625,
232
+ "rewards/chosen": -0.5947624444961548,
233
+ "rewards/margins": 0.22826068103313446,
234
+ "rewards/rejected": -0.8230231404304504,
235
+ "step": 130
236
+ },
237
+ {
238
+ "epoch": 0.24,
239
+ "grad_norm": 8.82315581796221,
240
+ "learning_rate": 4.992777627018238e-07,
241
+ "logits/chosen": -2.2996668815612793,
242
+ "logits/rejected": -2.334745168685913,
243
+ "logps/chosen": -287.4745788574219,
244
+ "logps/rejected": -276.83599853515625,
245
+ "loss": 0.6759,
246
+ "rewards/accuracies": 0.5625,
247
+ "rewards/chosen": -0.6079668402671814,
248
+ "rewards/margins": 0.3260408937931061,
249
+ "rewards/rejected": -0.9340078234672546,
250
+ "step": 140
251
+ },
252
+ {
253
+ "epoch": 0.26,
254
+ "grad_norm": 7.961709787250932,
255
+ "learning_rate": 4.985850694813964e-07,
256
+ "logits/chosen": -2.302722454071045,
257
+ "logits/rejected": -2.3677978515625,
258
+ "logps/chosen": -314.6284484863281,
259
+ "logps/rejected": -298.97955322265625,
260
+ "loss": 0.658,
261
+ "rewards/accuracies": 0.65625,
262
+ "rewards/chosen": -0.638712465763092,
263
+ "rewards/margins": 0.39987772703170776,
264
+ "rewards/rejected": -1.0385901927947998,
265
+ "step": 150
266
+ },
267
+ {
268
+ "epoch": 0.28,
269
+ "grad_norm": 8.327864420359786,
270
+ "learning_rate": 4.976624753622489e-07,
271
+ "logits/chosen": -2.3713505268096924,
272
+ "logits/rejected": -2.3584446907043457,
273
+ "logps/chosen": -329.738037109375,
274
+ "logps/rejected": -329.0250244140625,
275
+ "loss": 0.6513,
276
+ "rewards/accuracies": 0.6812499761581421,
277
+ "rewards/chosen": -0.5834181308746338,
278
+ "rewards/margins": 0.6477876901626587,
279
+ "rewards/rejected": -1.231205940246582,
280
+ "step": 160
281
+ },
282
+ {
283
+ "epoch": 0.3,
284
+ "grad_norm": 8.149557814002582,
285
+ "learning_rate": 4.965108335944079e-07,
286
+ "logits/chosen": -2.3453166484832764,
287
+ "logits/rejected": -2.3316988945007324,
288
+ "logps/chosen": -295.4629821777344,
289
+ "logps/rejected": -303.3232116699219,
290
+ "loss": 0.6467,
291
+ "rewards/accuracies": 0.637499988079071,
292
+ "rewards/chosen": -0.6690499782562256,
293
+ "rewards/margins": 0.5605732798576355,
294
+ "rewards/rejected": -1.2296231985092163,
295
+ "step": 170
296
+ },
297
+ {
298
+ "epoch": 0.31,
299
+ "grad_norm": 10.113037082589418,
300
+ "learning_rate": 4.951312092598509e-07,
301
+ "logits/chosen": -2.3545076847076416,
302
+ "logits/rejected": -2.3892111778259277,
303
+ "logps/chosen": -293.5394592285156,
304
+ "logps/rejected": -310.970947265625,
305
+ "loss": 0.6429,
306
+ "rewards/accuracies": 0.5874999761581421,
307
+ "rewards/chosen": -0.7175970077514648,
308
+ "rewards/margins": 0.3378085494041443,
309
+ "rewards/rejected": -1.0554054975509644,
310
+ "step": 180
311
+ },
312
+ {
313
+ "epoch": 0.33,
314
+ "grad_norm": 7.2425787929751975,
315
+ "learning_rate": 4.935248782874788e-07,
316
+ "logits/chosen": -2.3394787311553955,
317
+ "logits/rejected": -2.4027209281921387,
318
+ "logps/chosen": -322.01141357421875,
319
+ "logps/rejected": -312.63641357421875,
320
+ "loss": 0.6074,
321
+ "rewards/accuracies": 0.6812499761581421,
322
+ "rewards/chosen": -0.685249388217926,
323
+ "rewards/margins": 0.5340052843093872,
324
+ "rewards/rejected": -1.219254732131958,
325
+ "step": 190
326
+ },
327
+ {
328
+ "epoch": 0.35,
329
+ "grad_norm": 8.490660520819826,
330
+ "learning_rate": 4.916933262730875e-07,
331
+ "logits/chosen": -2.3357110023498535,
332
+ "logits/rejected": -2.3726134300231934,
333
+ "logps/chosen": -283.4224548339844,
334
+ "logps/rejected": -292.6156311035156,
335
+ "loss": 0.6273,
336
+ "rewards/accuracies": 0.625,
337
+ "rewards/chosen": -0.7353588342666626,
338
+ "rewards/margins": 0.36572274565696716,
339
+ "rewards/rejected": -1.1010816097259521,
340
+ "step": 200
341
+ },
342
+ {
343
+ "epoch": 0.35,
344
+ "eval_logits/chosen": -2.427767753601074,
345
+ "eval_logits/rejected": -2.5280518531799316,
346
+ "eval_logps/chosen": -305.2568054199219,
347
+ "eval_logps/rejected": -285.6261291503906,
348
+ "eval_loss": 0.5159560441970825,
349
+ "eval_rewards/accuracies": 0.6930379867553711,
350
+ "eval_rewards/chosen": -0.39940303564071655,
351
+ "eval_rewards/margins": 0.5467095971107483,
352
+ "eval_rewards/rejected": -0.9461126327514648,
353
+ "eval_runtime": 117.3962,
354
+ "eval_samples_per_second": 21.295,
355
+ "eval_steps_per_second": 0.673,
356
+ "step": 200
357
+ },
358
+ {
359
+ "epoch": 0.37,
360
+ "grad_norm": 8.576550556267508,
361
+ "learning_rate": 4.896382471054332e-07,
362
+ "logits/chosen": -2.3324050903320312,
363
+ "logits/rejected": -2.3912084102630615,
364
+ "logps/chosen": -288.8896789550781,
365
+ "logps/rejected": -268.7303161621094,
366
+ "loss": 0.6128,
367
+ "rewards/accuracies": 0.625,
368
+ "rewards/chosen": -0.6878677606582642,
369
+ "rewards/margins": 0.49019718170166016,
370
+ "rewards/rejected": -1.1780649423599243,
371
+ "step": 210
372
+ },
373
+ {
374
+ "epoch": 0.38,
375
+ "grad_norm": 7.314590858955313,
376
+ "learning_rate": 4.873615413996599e-07,
377
+ "logits/chosen": -2.3278536796569824,
378
+ "logits/rejected": -2.3542141914367676,
379
+ "logps/chosen": -320.20867919921875,
380
+ "logps/rejected": -344.2962646484375,
381
+ "loss": 0.6034,
382
+ "rewards/accuracies": 0.6499999761581421,
383
+ "rewards/chosen": -0.8140638470649719,
384
+ "rewards/margins": 0.6731799244880676,
385
+ "rewards/rejected": -1.487243890762329,
386
+ "step": 220
387
+ },
388
+ {
389
+ "epoch": 0.4,
390
+ "grad_norm": 7.89990023398186,
391
+ "learning_rate": 4.848653147395379e-07,
392
+ "logits/chosen": -2.3761696815490723,
393
+ "logits/rejected": -2.3647568225860596,
394
+ "logps/chosen": -278.48651123046875,
395
+ "logps/rejected": -313.1923522949219,
396
+ "loss": 0.5758,
397
+ "rewards/accuracies": 0.675000011920929,
398
+ "rewards/chosen": -0.8184002637863159,
399
+ "rewards/margins": 0.5446540117263794,
400
+ "rewards/rejected": -1.3630542755126953,
401
+ "step": 230
402
+ },
403
+ {
404
+ "epoch": 0.42,
405
+ "grad_norm": 6.865979693460653,
406
+ "learning_rate": 4.821518757301406e-07,
407
+ "logits/chosen": -2.3292605876922607,
408
+ "logits/rejected": -2.3688724040985107,
409
+ "logps/chosen": -266.1036071777344,
410
+ "logps/rejected": -297.202392578125,
411
+ "loss": 0.5331,
412
+ "rewards/accuracies": 0.7437499761581421,
413
+ "rewards/chosen": -0.7772084474563599,
414
+ "rewards/margins": 0.8323043584823608,
415
+ "rewards/rejected": -1.6095129251480103,
416
+ "step": 240
417
+ },
418
+ {
419
+ "epoch": 0.44,
420
+ "grad_norm": 7.1462197132774925,
421
+ "learning_rate": 4.792237338627589e-07,
422
+ "logits/chosen": -2.360563278198242,
423
+ "logits/rejected": -2.3513259887695312,
424
+ "logps/chosen": -300.3686828613281,
425
+ "logps/rejected": -312.9150390625,
426
+ "loss": 0.5817,
427
+ "rewards/accuracies": 0.6812499761581421,
428
+ "rewards/chosen": -0.8266876339912415,
429
+ "rewards/margins": 0.8314968347549438,
430
+ "rewards/rejected": -1.6581846475601196,
431
+ "step": 250
432
+ },
433
+ {
434
+ "epoch": 0.45,
435
+ "grad_norm": 7.9450903494413865,
436
+ "learning_rate": 4.76083597194028e-07,
437
+ "logits/chosen": -2.3549387454986572,
438
+ "logits/rejected": -2.3544278144836426,
439
+ "logps/chosen": -274.65362548828125,
440
+ "logps/rejected": -294.8714294433594,
441
+ "loss": 0.5582,
442
+ "rewards/accuracies": 0.6312500238418579,
443
+ "rewards/chosen": -0.8873316645622253,
444
+ "rewards/margins": 0.63465416431427,
445
+ "rewards/rejected": -1.5219857692718506,
446
+ "step": 260
447
+ },
448
+ {
449
+ "epoch": 0.47,
450
+ "grad_norm": 8.86333832366373,
451
+ "learning_rate": 4.72734369841415e-07,
452
+ "logits/chosen": -2.268969774246216,
453
+ "logits/rejected": -2.290459632873535,
454
+ "logps/chosen": -326.5730285644531,
455
+ "logps/rejected": -340.42547607421875,
456
+ "loss": 0.5535,
457
+ "rewards/accuracies": 0.6312500238418579,
458
+ "rewards/chosen": -0.9808245897293091,
459
+ "rewards/margins": 0.7018911838531494,
460
+ "rewards/rejected": -1.6827157735824585,
461
+ "step": 270
462
+ },
463
+ {
464
+ "epoch": 0.49,
465
+ "grad_norm": 9.058341702180591,
466
+ "learning_rate": 4.691791492973796e-07,
467
+ "logits/chosen": -2.2908778190612793,
468
+ "logits/rejected": -2.3170814514160156,
469
+ "logps/chosen": -285.4623107910156,
470
+ "logps/rejected": -287.90460205078125,
471
+ "loss": 0.568,
472
+ "rewards/accuracies": 0.7124999761581421,
473
+ "rewards/chosen": -0.9195302128791809,
474
+ "rewards/margins": 0.8237813115119934,
475
+ "rewards/rejected": -1.7433115243911743,
476
+ "step": 280
477
+ },
478
+ {
479
+ "epoch": 0.51,
480
+ "grad_norm": 8.549379026648278,
481
+ "learning_rate": 4.6542122356469673e-07,
482
+ "logits/chosen": -2.3191022872924805,
483
+ "logits/rejected": -2.2784862518310547,
484
+ "logps/chosen": -290.696533203125,
485
+ "logps/rejected": -345.26922607421875,
486
+ "loss": 0.566,
487
+ "rewards/accuracies": 0.7250000238418579,
488
+ "rewards/chosen": -0.867855429649353,
489
+ "rewards/margins": 1.0541679859161377,
490
+ "rewards/rejected": -1.9220234155654907,
491
+ "step": 290
492
+ },
493
+ {
494
+ "epoch": 0.52,
495
+ "grad_norm": 8.117986354757171,
496
+ "learning_rate": 4.6146406811558627e-07,
497
+ "logits/chosen": -2.2675626277923584,
498
+ "logits/rejected": -2.2998242378234863,
499
+ "logps/chosen": -300.87139892578125,
500
+ "logps/rejected": -308.83258056640625,
501
+ "loss": 0.5538,
502
+ "rewards/accuracies": 0.668749988079071,
503
+ "rewards/chosen": -1.039670467376709,
504
+ "rewards/margins": 0.7833054065704346,
505
+ "rewards/rejected": -1.8229758739471436,
506
+ "step": 300
507
+ },
508
+ {
509
+ "epoch": 0.52,
510
+ "eval_logits/chosen": -2.410975217819214,
511
+ "eval_logits/rejected": -2.4996373653411865,
512
+ "eval_logps/chosen": -310.4469909667969,
513
+ "eval_logps/rejected": -298.4869689941406,
514
+ "eval_loss": 0.47806617617607117,
515
+ "eval_rewards/accuracies": 0.7246835231781006,
516
+ "eval_rewards/chosen": -0.6589112877845764,
517
+ "eval_rewards/margins": 0.9302425980567932,
518
+ "eval_rewards/rejected": -1.58915376663208,
519
+ "eval_runtime": 116.9313,
520
+ "eval_samples_per_second": 21.38,
521
+ "eval_steps_per_second": 0.676,
522
+ "step": 300
523
+ },
524
+ {
525
+ "epoch": 0.54,
526
+ "grad_norm": 7.789987130313136,
527
+ "learning_rate": 4.5731134267746464e-07,
528
+ "logits/chosen": -2.3223114013671875,
529
+ "logits/rejected": -2.351526975631714,
530
+ "logps/chosen": -294.28973388671875,
531
+ "logps/rejected": -294.4921569824219,
532
+ "loss": 0.5404,
533
+ "rewards/accuracies": 0.6875,
534
+ "rewards/chosen": -0.9947312474250793,
535
+ "rewards/margins": 0.8857283592224121,
536
+ "rewards/rejected": -1.8804595470428467,
537
+ "step": 310
538
+ },
539
+ {
540
+ "epoch": 0.56,
541
+ "grad_norm": 6.695877210721463,
542
+ "learning_rate": 4.529668878482905e-07,
543
+ "logits/chosen": -2.278529644012451,
544
+ "logits/rejected": -2.322715997695923,
545
+ "logps/chosen": -302.17156982421875,
546
+ "logps/rejected": -311.0806884765625,
547
+ "loss": 0.5469,
548
+ "rewards/accuracies": 0.699999988079071,
549
+ "rewards/chosen": -0.9337432980537415,
550
+ "rewards/margins": 0.9280776977539062,
551
+ "rewards/rejected": -1.861820936203003,
552
+ "step": 320
553
+ },
554
+ {
555
+ "epoch": 0.57,
556
+ "grad_norm": 7.390355171835368,
557
+ "learning_rate": 4.484347215446336e-07,
558
+ "logits/chosen": -2.2755682468414307,
559
+ "logits/rejected": -2.3354978561401367,
560
+ "logps/chosen": -309.1221008300781,
561
+ "logps/rejected": -321.5992431640625,
562
+ "loss": 0.5314,
563
+ "rewards/accuracies": 0.7437499761581421,
564
+ "rewards/chosen": -0.9960958361625671,
565
+ "rewards/margins": 0.8858348727226257,
566
+ "rewards/rejected": -1.8819307088851929,
567
+ "step": 330
568
+ },
569
+ {
570
+ "epoch": 0.59,
571
+ "grad_norm": 7.988169852743586,
572
+ "learning_rate": 4.4371903528575345e-07,
573
+ "logits/chosen": -2.279055118560791,
574
+ "logits/rejected": -2.293339490890503,
575
+ "logps/chosen": -313.377685546875,
576
+ "logps/rejected": -320.6536560058594,
577
+ "loss": 0.5217,
578
+ "rewards/accuracies": 0.7250000238418579,
579
+ "rewards/chosen": -1.1063129901885986,
580
+ "rewards/margins": 0.950600802898407,
581
+ "rewards/rejected": -2.0569138526916504,
582
+ "step": 340
583
+ },
584
+ {
585
+ "epoch": 0.61,
586
+ "grad_norm": 8.25715266845146,
587
+ "learning_rate": 4.3882419031712327e-07,
588
+ "logits/chosen": -2.3251774311065674,
589
+ "logits/rejected": -2.3021457195281982,
590
+ "logps/chosen": -284.44891357421875,
591
+ "logps/rejected": -304.513671875,
592
+ "loss": 0.5226,
593
+ "rewards/accuracies": 0.6812499761581421,
594
+ "rewards/chosen": -0.952631950378418,
595
+ "rewards/margins": 1.1419737339019775,
596
+ "rewards/rejected": -2.0946059226989746,
597
+ "step": 350
598
+ },
599
+ {
600
+ "epoch": 0.63,
601
+ "grad_norm": 7.634851788477121,
602
+ "learning_rate": 4.33754713576985e-07,
603
+ "logits/chosen": -2.2681329250335693,
604
+ "logits/rejected": -2.3093745708465576,
605
+ "logps/chosen": -264.74871826171875,
606
+ "logps/rejected": -277.2441711425781,
607
+ "loss": 0.554,
608
+ "rewards/accuracies": 0.625,
609
+ "rewards/chosen": -1.1214876174926758,
610
+ "rewards/margins": 0.9260104298591614,
611
+ "rewards/rejected": -2.0474982261657715,
612
+ "step": 360
613
+ },
614
+ {
615
+ "epoch": 0.64,
616
+ "grad_norm": 7.279488194032858,
617
+ "learning_rate": 4.2851529350966486e-07,
618
+ "logits/chosen": -2.2603583335876465,
619
+ "logits/rejected": -2.266740560531616,
620
+ "logps/chosen": -283.461669921875,
621
+ "logps/rejected": -318.08355712890625,
622
+ "loss": 0.5216,
623
+ "rewards/accuracies": 0.731249988079071,
624
+ "rewards/chosen": -1.002974271774292,
625
+ "rewards/margins": 1.0799973011016846,
626
+ "rewards/rejected": -2.0829715728759766,
627
+ "step": 370
628
+ },
629
+ {
630
+ "epoch": 0.66,
631
+ "grad_norm": 7.8421017282099585,
632
+ "learning_rate": 4.231107757295223e-07,
633
+ "logits/chosen": -2.2358787059783936,
634
+ "logits/rejected": -2.279066801071167,
635
+ "logps/chosen": -297.420654296875,
636
+ "logps/rejected": -306.3590393066406,
637
+ "loss": 0.5241,
638
+ "rewards/accuracies": 0.706250011920929,
639
+ "rewards/chosen": -1.1632144451141357,
640
+ "rewards/margins": 1.0475361347198486,
641
+ "rewards/rejected": -2.2107505798339844,
642
+ "step": 380
643
+ },
644
+ {
645
+ "epoch": 0.68,
646
+ "grad_norm": 7.394520753535137,
647
+ "learning_rate": 4.175461585395423e-07,
648
+ "logits/chosen": -2.217916965484619,
649
+ "logits/rejected": -2.3138163089752197,
650
+ "logps/chosen": -317.17303466796875,
651
+ "logps/rejected": -288.50323486328125,
652
+ "loss": 0.5381,
653
+ "rewards/accuracies": 0.7562500238418579,
654
+ "rewards/chosen": -0.9419029951095581,
655
+ "rewards/margins": 1.1881153583526611,
656
+ "rewards/rejected": -2.130018472671509,
657
+ "step": 390
658
+ },
659
+ {
660
+ "epoch": 0.7,
661
+ "grad_norm": 8.212800923068182,
662
+ "learning_rate": 4.118265883087153e-07,
663
+ "logits/chosen": -2.2654032707214355,
664
+ "logits/rejected": -2.2785887718200684,
665
+ "logps/chosen": -285.03448486328125,
666
+ "logps/rejected": -328.69390869140625,
667
+ "loss": 0.5056,
668
+ "rewards/accuracies": 0.7437499761581421,
669
+ "rewards/chosen": -1.0962527990341187,
670
+ "rewards/margins": 1.4823755025863647,
671
+ "rewards/rejected": -2.5786283016204834,
672
+ "step": 400
673
+ },
674
+ {
675
+ "epoch": 0.7,
676
+ "eval_logits/chosen": -2.364351511001587,
677
+ "eval_logits/rejected": -2.447211265563965,
678
+ "eval_logps/chosen": -313.8343505859375,
679
+ "eval_logps/rejected": -309.3686828613281,
680
+ "eval_loss": 0.45939481258392334,
681
+ "eval_rewards/accuracies": 0.7436708807945251,
682
+ "eval_rewards/chosen": -0.8282797932624817,
683
+ "eval_rewards/margins": 1.3049596548080444,
684
+ "eval_rewards/rejected": -2.133239507675171,
685
+ "eval_runtime": 116.733,
686
+ "eval_samples_per_second": 21.416,
687
+ "eval_steps_per_second": 0.677,
688
+ "step": 400
689
+ },
690
+ {
691
+ "epoch": 0.71,
692
+ "grad_norm": 8.827849836435306,
693
+ "learning_rate": 4.059573547124793e-07,
694
+ "logits/chosen": -2.2229080200195312,
695
+ "logits/rejected": -2.2402656078338623,
696
+ "logps/chosen": -290.5343933105469,
697
+ "logps/rejected": -333.4751892089844,
698
+ "loss": 0.5073,
699
+ "rewards/accuracies": 0.731249988079071,
700
+ "rewards/chosen": -1.0698788166046143,
701
+ "rewards/margins": 1.2725006341934204,
702
+ "rewards/rejected": -2.342379331588745,
703
+ "step": 410
704
+ },
705
+ {
706
+ "epoch": 0.73,
707
+ "grad_norm": 17.168544254244743,
708
+ "learning_rate": 3.9994388584062835e-07,
709
+ "logits/chosen": -2.2369210720062256,
710
+ "logits/rejected": -2.2701995372772217,
711
+ "logps/chosen": -310.66650390625,
712
+ "logps/rejected": -343.23736572265625,
713
+ "loss": 0.4973,
714
+ "rewards/accuracies": 0.737500011920929,
715
+ "rewards/chosen": -1.10798180103302,
716
+ "rewards/margins": 1.2380707263946533,
717
+ "rewards/rejected": -2.346052646636963,
718
+ "step": 420
719
+ },
720
+ {
721
+ "epoch": 0.75,
722
+ "grad_norm": 7.900686275091842,
723
+ "learning_rate": 3.9379174317720867e-07,
724
+ "logits/chosen": -2.2615458965301514,
725
+ "logits/rejected": -2.2584452629089355,
726
+ "logps/chosen": -253.0293731689453,
727
+ "logps/rejected": -310.45562744140625,
728
+ "loss": 0.5226,
729
+ "rewards/accuracies": 0.731249988079071,
730
+ "rewards/chosen": -1.0362132787704468,
731
+ "rewards/margins": 1.1289708614349365,
732
+ "rewards/rejected": -2.1651840209960938,
733
+ "step": 430
734
+ },
735
+ {
736
+ "epoch": 0.77,
737
+ "grad_norm": 7.866159803847311,
738
+ "learning_rate": 3.875066164570476e-07,
739
+ "logits/chosen": -2.2734320163726807,
740
+ "logits/rejected": -2.2685108184814453,
741
+ "logps/chosen": -310.897705078125,
742
+ "logps/rejected": -311.92559814453125,
743
+ "loss": 0.4877,
744
+ "rewards/accuracies": 0.6937500238418579,
745
+ "rewards/chosen": -0.9617986679077148,
746
+ "rewards/margins": 1.4187655448913574,
747
+ "rewards/rejected": -2.3805642127990723,
748
+ "step": 440
749
+ },
750
+ {
751
+ "epoch": 0.78,
752
+ "grad_norm": 8.068518521527759,
753
+ "learning_rate": 3.8109431840367094e-07,
754
+ "logits/chosen": -2.2477684020996094,
755
+ "logits/rejected": -2.2789039611816406,
756
+ "logps/chosen": -315.4168395996094,
757
+ "logps/rejected": -322.646728515625,
758
+ "loss": 0.5117,
759
+ "rewards/accuracies": 0.762499988079071,
760
+ "rewards/chosen": -1.0208594799041748,
761
+ "rewards/margins": 1.3095576763153076,
762
+ "rewards/rejected": -2.3304171562194824,
763
+ "step": 450
764
+ },
765
+ {
766
+ "epoch": 0.8,
767
+ "grad_norm": 8.68774893484682,
768
+ "learning_rate": 3.7456077935347586e-07,
769
+ "logits/chosen": -2.279040813446045,
770
+ "logits/rejected": -2.3296942710876465,
771
+ "logps/chosen": -268.678955078125,
772
+ "logps/rejected": -305.7779235839844,
773
+ "loss": 0.5259,
774
+ "rewards/accuracies": 0.7562500238418579,
775
+ "rewards/chosen": -0.930595874786377,
776
+ "rewards/margins": 1.5479636192321777,
777
+ "rewards/rejected": -2.4785592555999756,
778
+ "step": 460
779
+ },
780
+ {
781
+ "epoch": 0.82,
782
+ "grad_norm": 7.377326318454655,
783
+ "learning_rate": 3.6791204177113076e-07,
784
+ "logits/chosen": -2.2461414337158203,
785
+ "logits/rejected": -2.261500358581543,
786
+ "logps/chosen": -288.28131103515625,
787
+ "logps/rejected": -323.84259033203125,
788
+ "loss": 0.5055,
789
+ "rewards/accuracies": 0.699999988079071,
790
+ "rewards/chosen": -0.9557926058769226,
791
+ "rewards/margins": 1.3024319410324097,
792
+ "rewards/rejected": -2.2582249641418457,
793
+ "step": 470
794
+ },
795
+ {
796
+ "epoch": 0.84,
797
+ "grad_norm": 9.031552997233952,
798
+ "learning_rate": 3.6115425466127523e-07,
799
+ "logits/chosen": -2.211353302001953,
800
+ "logits/rejected": -2.285994529724121,
801
+ "logps/chosen": -307.39508056640625,
802
+ "logps/rejected": -309.01617431640625,
803
+ "loss": 0.5169,
804
+ "rewards/accuracies": 0.706250011920929,
805
+ "rewards/chosen": -1.0073463916778564,
806
+ "rewards/margins": 1.2757568359375,
807
+ "rewards/rejected": -2.2831032276153564,
808
+ "step": 480
809
+ },
810
+ {
811
+ "epoch": 0.85,
812
+ "grad_norm": 8.363906043420153,
813
+ "learning_rate": 3.5429366788168686e-07,
814
+ "logits/chosen": -2.2397665977478027,
815
+ "logits/rejected": -2.2399213314056396,
816
+ "logps/chosen": -320.8814697265625,
817
+ "logps/rejected": -316.41204833984375,
818
+ "loss": 0.5137,
819
+ "rewards/accuracies": 0.6812499761581421,
820
+ "rewards/chosen": -1.077854871749878,
821
+ "rewards/margins": 1.0860309600830078,
822
+ "rewards/rejected": -2.1638855934143066,
823
+ "step": 490
824
+ },
825
+ {
826
+ "epoch": 0.87,
827
+ "grad_norm": 8.777454709865056,
828
+ "learning_rate": 3.4733662636317615e-07,
829
+ "logits/chosen": -2.26035213470459,
830
+ "logits/rejected": -2.282977342605591,
831
+ "logps/chosen": -323.86785888671875,
832
+ "logps/rejected": -323.78558349609375,
833
+ "loss": 0.4983,
834
+ "rewards/accuracies": 0.7437499761581421,
835
+ "rewards/chosen": -1.0498888492584229,
836
+ "rewards/margins": 1.2481728792190552,
837
+ "rewards/rejected": -2.2980618476867676,
838
+ "step": 500
839
+ },
840
+ {
841
+ "epoch": 0.87,
842
+ "eval_logits/chosen": -2.3404455184936523,
843
+ "eval_logits/rejected": -2.422346830368042,
844
+ "eval_logps/chosen": -312.7843322753906,
845
+ "eval_logps/rejected": -312.31671142578125,
846
+ "eval_loss": 0.4511754512786865,
847
+ "eval_rewards/accuracies": 0.746835470199585,
848
+ "eval_rewards/chosen": -0.7757795453071594,
849
+ "eval_rewards/margins": 1.5048617124557495,
850
+ "eval_rewards/rejected": -2.2806413173675537,
851
+ "eval_runtime": 116.6502,
852
+ "eval_samples_per_second": 21.432,
853
+ "eval_steps_per_second": 0.677,
854
+ "step": 500
855
+ },
856
+ {
857
+ "epoch": 0.89,
858
+ "grad_norm": 7.734353981694666,
859
+ "learning_rate": 3.4028956424155383e-07,
860
+ "logits/chosen": -2.2213361263275146,
861
+ "logits/rejected": -2.2351489067077637,
862
+ "logps/chosen": -306.02081298828125,
863
+ "logps/rejected": -336.88385009765625,
864
+ "loss": 0.489,
865
+ "rewards/accuracies": 0.7250000238418579,
866
+ "rewards/chosen": -1.0291322469711304,
867
+ "rewards/margins": 1.3191864490509033,
868
+ "rewards/rejected": -2.3483192920684814,
869
+ "step": 510
870
+ },
871
+ {
872
+ "epoch": 0.91,
873
+ "grad_norm": 8.664121177886866,
874
+ "learning_rate": 3.3315899890709766e-07,
875
+ "logits/chosen": -2.2391321659088135,
876
+ "logits/rejected": -2.2733283042907715,
877
+ "logps/chosen": -320.98895263671875,
878
+ "logps/rejected": -328.81097412109375,
879
+ "loss": 0.489,
880
+ "rewards/accuracies": 0.731249988079071,
881
+ "rewards/chosen": -1.1005600690841675,
882
+ "rewards/margins": 1.4874314069747925,
883
+ "rewards/rejected": -2.58799147605896,
884
+ "step": 520
885
+ },
886
+ {
887
+ "epoch": 0.92,
888
+ "grad_norm": 8.147807181233981,
889
+ "learning_rate": 3.259515249770236e-07,
890
+ "logits/chosen": -2.1741220951080322,
891
+ "logits/rejected": -2.2048075199127197,
892
+ "logps/chosen": -299.4497375488281,
893
+ "logps/rejected": -341.54168701171875,
894
+ "loss": 0.5198,
895
+ "rewards/accuracies": 0.800000011920929,
896
+ "rewards/chosen": -1.1551388502120972,
897
+ "rewards/margins": 1.6270313262939453,
898
+ "rewards/rejected": -2.782170057296753,
899
+ "step": 530
900
+ },
901
+ {
902
+ "epoch": 0.94,
903
+ "grad_norm": 7.750929958584628,
904
+ "learning_rate": 3.186738081965329e-07,
905
+ "logits/chosen": -2.202561616897583,
906
+ "logits/rejected": -2.2528486251831055,
907
+ "logps/chosen": -294.01123046875,
908
+ "logps/rejected": -289.1777648925781,
909
+ "loss": 0.5261,
910
+ "rewards/accuracies": 0.731249988079071,
911
+ "rewards/chosen": -1.140953779220581,
912
+ "rewards/margins": 1.2280247211456299,
913
+ "rewards/rejected": -2.368978500366211,
914
+ "step": 540
915
+ },
916
+ {
917
+ "epoch": 0.96,
918
+ "grad_norm": 11.076352034341378,
919
+ "learning_rate": 3.1133257927407875e-07,
920
+ "logits/chosen": -2.1735188961029053,
921
+ "logits/rejected": -2.2224655151367188,
922
+ "logps/chosen": -305.57177734375,
923
+ "logps/rejected": -329.1368713378906,
924
+ "loss": 0.479,
925
+ "rewards/accuracies": 0.6875,
926
+ "rewards/chosen": -1.2371329069137573,
927
+ "rewards/margins": 1.3205959796905518,
928
+ "rewards/rejected": -2.5577290058135986,
929
+ "step": 550
930
+ },
931
+ {
932
+ "epoch": 0.98,
933
+ "grad_norm": 8.548263020458736,
934
+ "learning_rate": 3.0393462765655133e-07,
935
+ "logits/chosen": -2.256953001022339,
936
+ "logits/rejected": -2.2490811347961426,
937
+ "logps/chosen": -324.26397705078125,
938
+ "logps/rejected": -332.2065124511719,
939
+ "loss": 0.5242,
940
+ "rewards/accuracies": 0.706250011920929,
941
+ "rewards/chosen": -1.207909345626831,
942
+ "rewards/margins": 1.1256216764450073,
943
+ "rewards/rejected": -2.333531141281128,
944
+ "step": 560
945
+ },
946
+ {
947
+ "epoch": 0.99,
948
+ "grad_norm": 10.129228469440344,
949
+ "learning_rate": 2.9648679525014047e-07,
950
+ "logits/chosen": -2.249263286590576,
951
+ "logits/rejected": -2.224191188812256,
952
+ "logps/chosen": -301.6817626953125,
953
+ "logps/rejected": -317.6774597167969,
954
+ "loss": 0.5166,
955
+ "rewards/accuracies": 0.78125,
956
+ "rewards/chosen": -0.9407655000686646,
957
+ "rewards/margins": 1.2918128967285156,
958
+ "rewards/rejected": -2.2325782775878906,
959
+ "step": 570
960
+ },
961
+ {
962
+ "epoch": 1.01,
963
+ "grad_norm": 7.601401935918346,
964
+ "learning_rate": 2.88995970092681e-07,
965
+ "logits/chosen": -2.228137254714966,
966
+ "logits/rejected": -2.2642438411712646,
967
+ "logps/chosen": -346.5960388183594,
968
+ "logps/rejected": -334.5989074707031,
969
+ "loss": 0.4668,
970
+ "rewards/accuracies": 0.793749988079071,
971
+ "rewards/chosen": -1.0293288230895996,
972
+ "rewards/margins": 1.4567983150482178,
973
+ "rewards/rejected": -2.4861273765563965,
974
+ "step": 580
975
+ },
976
+ {
977
+ "epoch": 1.03,
978
+ "grad_norm": 7.059735053355466,
979
+ "learning_rate": 2.8146907998333414e-07,
980
+ "logits/chosen": -2.2480270862579346,
981
+ "logits/rejected": -2.2125768661499023,
982
+ "logps/chosen": -292.4286804199219,
983
+ "logps/rejected": -345.6949157714844,
984
+ "loss": 0.4438,
985
+ "rewards/accuracies": 0.7875000238418579,
986
+ "rewards/chosen": -0.9727999567985535,
987
+ "rewards/margins": 1.7654082775115967,
988
+ "rewards/rejected": -2.738208293914795,
989
+ "step": 590
990
+ },
991
+ {
992
+ "epoch": 1.04,
993
+ "grad_norm": 7.162224042755479,
994
+ "learning_rate": 2.7391308607549617e-07,
995
+ "logits/chosen": -2.2136716842651367,
996
+ "logits/rejected": -2.257483720779419,
997
+ "logps/chosen": -302.130126953125,
998
+ "logps/rejected": -333.96771240234375,
999
+ "loss": 0.4662,
1000
+ "rewards/accuracies": 0.737500011920929,
1001
+ "rewards/chosen": -1.0814803838729858,
1002
+ "rewards/margins": 1.416599988937378,
1003
+ "rewards/rejected": -2.498080253601074,
1004
+ "step": 600
1005
+ },
1006
+ {
1007
+ "epoch": 1.04,
1008
+ "eval_logits/chosen": -2.3215160369873047,
1009
+ "eval_logits/rejected": -2.4048969745635986,
1010
+ "eval_logps/chosen": -312.94647216796875,
1011
+ "eval_logps/rejected": -314.7354736328125,
1012
+ "eval_loss": 0.44313427805900574,
1013
+ "eval_rewards/accuracies": 0.7658227682113647,
1014
+ "eval_rewards/chosen": -0.78388512134552,
1015
+ "eval_rewards/margins": 1.6176937818527222,
1016
+ "eval_rewards/rejected": -2.401578903198242,
1017
+ "eval_runtime": 116.9664,
1018
+ "eval_samples_per_second": 21.374,
1019
+ "eval_steps_per_second": 0.675,
1020
+ "step": 600
1021
+ },
1022
+ {
1023
+ "epoch": 1.06,
1024
+ "grad_norm": 5.7685048393945015,
1025
+ "learning_rate": 2.663349764388602e-07,
1026
+ "logits/chosen": -2.224734306335449,
1027
+ "logits/rejected": -2.222229480743408,
1028
+ "logps/chosen": -305.4608154296875,
1029
+ "logps/rejected": -343.47027587890625,
1030
+ "loss": 0.4284,
1031
+ "rewards/accuracies": 0.8125,
1032
+ "rewards/chosen": -0.9569365382194519,
1033
+ "rewards/margins": 1.9764817953109741,
1034
+ "rewards/rejected": -2.9334182739257812,
1035
+ "step": 610
1036
+ },
1037
+ {
1038
+ "epoch": 1.08,
1039
+ "grad_norm": 7.994083679128179,
1040
+ "learning_rate": 2.587417595965833e-07,
1041
+ "logits/chosen": -2.2422053813934326,
1042
+ "logits/rejected": -2.2728614807128906,
1043
+ "logps/chosen": -274.6269226074219,
1044
+ "logps/rejected": -284.4000549316406,
1045
+ "loss": 0.4676,
1046
+ "rewards/accuracies": 0.75,
1047
+ "rewards/chosen": -0.9759398698806763,
1048
+ "rewards/margins": 1.389269232749939,
1049
+ "rewards/rejected": -2.3652091026306152,
1050
+ "step": 620
1051
+ },
1052
+ {
1053
+ "epoch": 1.1,
1054
+ "grad_norm": 6.234612520244718,
1055
+ "learning_rate": 2.511404580435399e-07,
1056
+ "logits/chosen": -2.2112672328948975,
1057
+ "logits/rejected": -2.222592353820801,
1058
+ "logps/chosen": -300.41253662109375,
1059
+ "logps/rejected": -317.8959655761719,
1060
+ "loss": 0.4188,
1061
+ "rewards/accuracies": 0.768750011920929,
1062
+ "rewards/chosen": -1.077165961265564,
1063
+ "rewards/margins": 1.776564598083496,
1064
+ "rewards/rejected": -2.8537306785583496,
1065
+ "step": 630
1066
+ },
1067
+ {
1068
+ "epoch": 1.11,
1069
+ "grad_norm": 7.4969961974893184,
1070
+ "learning_rate": 2.435381017516511e-07,
1071
+ "logits/chosen": -2.182955026626587,
1072
+ "logits/rejected": -2.194115161895752,
1073
+ "logps/chosen": -302.7352600097656,
1074
+ "logps/rejected": -356.3200378417969,
1075
+ "loss": 0.4409,
1076
+ "rewards/accuracies": 0.737500011920929,
1077
+ "rewards/chosen": -1.1854764223098755,
1078
+ "rewards/margins": 1.6020820140838623,
1079
+ "rewards/rejected": -2.7875583171844482,
1080
+ "step": 640
1081
+ },
1082
+ {
1083
+ "epoch": 1.13,
1084
+ "grad_norm": 7.325319067879207,
1085
+ "learning_rate": 2.3594172166830066e-07,
1086
+ "logits/chosen": -2.1576247215270996,
1087
+ "logits/rejected": -2.212817907333374,
1088
+ "logps/chosen": -327.29693603515625,
1089
+ "logps/rejected": -323.61724853515625,
1090
+ "loss": 0.4553,
1091
+ "rewards/accuracies": 0.800000011920929,
1092
+ "rewards/chosen": -1.041488766670227,
1093
+ "rewards/margins": 1.4408646821975708,
1094
+ "rewards/rejected": -2.4823532104492188,
1095
+ "step": 650
1096
+ },
1097
+ {
1098
+ "epoch": 1.15,
1099
+ "grad_norm": 6.790105275708977,
1100
+ "learning_rate": 2.2835834321384799e-07,
1101
+ "logits/chosen": -2.214137554168701,
1102
+ "logits/rejected": -2.2264151573181152,
1103
+ "logps/chosen": -321.0436706542969,
1104
+ "logps/rejected": -328.35552978515625,
1105
+ "loss": 0.4503,
1106
+ "rewards/accuracies": 0.762499988079071,
1107
+ "rewards/chosen": -1.0654566287994385,
1108
+ "rewards/margins": 1.718808889389038,
1109
+ "rewards/rejected": -2.7842652797698975,
1110
+ "step": 660
1111
+ },
1112
+ {
1113
+ "epoch": 1.17,
1114
+ "grad_norm": 7.062324933750996,
1115
+ "learning_rate": 2.20794979784253e-07,
1116
+ "logits/chosen": -2.1624903678894043,
1117
+ "logits/rejected": -2.247901678085327,
1118
+ "logps/chosen": -319.1899719238281,
1119
+ "logps/rejected": -320.76129150390625,
1120
+ "loss": 0.4674,
1121
+ "rewards/accuracies": 0.7749999761581421,
1122
+ "rewards/chosen": -1.059831976890564,
1123
+ "rewards/margins": 1.5487343072891235,
1124
+ "rewards/rejected": -2.6085660457611084,
1125
+ "step": 670
1126
+ },
1127
+ {
1128
+ "epoch": 1.18,
1129
+ "grad_norm": 6.8992559586043845,
1130
+ "learning_rate": 2.132586262648217e-07,
1131
+ "logits/chosen": -2.1993870735168457,
1132
+ "logits/rejected": -2.2455170154571533,
1133
+ "logps/chosen": -342.77874755859375,
1134
+ "logps/rejected": -336.9671936035156,
1135
+ "loss": 0.4273,
1136
+ "rewards/accuracies": 0.75,
1137
+ "rewards/chosen": -1.1384742259979248,
1138
+ "rewards/margins": 1.6727256774902344,
1139
+ "rewards/rejected": -2.811199903488159,
1140
+ "step": 680
1141
+ },
1142
+ {
1143
+ "epoch": 1.2,
1144
+ "grad_norm": 7.798612367752025,
1145
+ "learning_rate": 2.0575625256107107e-07,
1146
+ "logits/chosen": -2.1738362312316895,
1147
+ "logits/rejected": -2.1713316440582275,
1148
+ "logps/chosen": -282.6869201660156,
1149
+ "logps/rejected": -328.921142578125,
1150
+ "loss": 0.4407,
1151
+ "rewards/accuracies": 0.793749988079071,
1152
+ "rewards/chosen": -0.9972604513168335,
1153
+ "rewards/margins": 1.8666757345199585,
1154
+ "rewards/rejected": -2.863936424255371,
1155
+ "step": 690
1156
+ },
1157
+ {
1158
+ "epoch": 1.22,
1159
+ "grad_norm": 7.992242941279979,
1160
+ "learning_rate": 1.9829479715269584e-07,
1161
+ "logits/chosen": -2.2045226097106934,
1162
+ "logits/rejected": -2.209587812423706,
1163
+ "logps/chosen": -309.096923828125,
1164
+ "logps/rejected": -335.237548828125,
1165
+ "loss": 0.4411,
1166
+ "rewards/accuracies": 0.7749999761581421,
1167
+ "rewards/chosen": -1.1609227657318115,
1168
+ "rewards/margins": 1.7247358560562134,
1169
+ "rewards/rejected": -2.8856582641601562,
1170
+ "step": 700
1171
+ },
1172
+ {
1173
+ "epoch": 1.22,
1174
+ "eval_logits/chosen": -2.3015921115875244,
1175
+ "eval_logits/rejected": -2.3840179443359375,
1176
+ "eval_logps/chosen": -317.4481201171875,
1177
+ "eval_logps/rejected": -321.8678894042969,
1178
+ "eval_loss": 0.44147032499313354,
1179
+ "eval_rewards/accuracies": 0.7689873576164246,
1180
+ "eval_rewards/chosen": -1.0089699029922485,
1181
+ "eval_rewards/margins": 1.7492305040359497,
1182
+ "eval_rewards/rejected": -2.7582004070281982,
1183
+ "eval_runtime": 116.9747,
1184
+ "eval_samples_per_second": 21.372,
1185
+ "eval_steps_per_second": 0.675,
1186
+ "step": 700
1187
+ },
1188
+ {
1189
+ "epoch": 1.24,
1190
+ "grad_norm": 8.274421110713899,
1191
+ "learning_rate": 1.908811606765996e-07,
1192
+ "logits/chosen": -2.184628963470459,
1193
+ "logits/rejected": -2.218219518661499,
1194
+ "logps/chosen": -313.20037841796875,
1195
+ "logps/rejected": -332.700927734375,
1196
+ "loss": 0.4436,
1197
+ "rewards/accuracies": 0.75,
1198
+ "rewards/chosen": -1.1558970212936401,
1199
+ "rewards/margins": 1.7350355386734009,
1200
+ "rewards/rejected": -2.89093279838562,
1201
+ "step": 710
1202
+ },
1203
+ {
1204
+ "epoch": 1.25,
1205
+ "grad_norm": 8.088194505594497,
1206
+ "learning_rate": 1.8352219954492414e-07,
1207
+ "logits/chosen": -2.1805386543273926,
1208
+ "logits/rejected": -2.221952438354492,
1209
+ "logps/chosen": -306.16717529296875,
1210
+ "logps/rejected": -312.39544677734375,
1211
+ "loss": 0.4605,
1212
+ "rewards/accuracies": 0.762499988079071,
1213
+ "rewards/chosen": -1.1549646854400635,
1214
+ "rewards/margins": 1.5584920644760132,
1215
+ "rewards/rejected": -2.713456630706787,
1216
+ "step": 720
1217
+ },
1218
+ {
1219
+ "epoch": 1.27,
1220
+ "grad_norm": 5.472862021460433,
1221
+ "learning_rate": 1.7622471960397922e-07,
1222
+ "logits/chosen": -2.1999268531799316,
1223
+ "logits/rejected": -2.2310051918029785,
1224
+ "logps/chosen": -297.07659912109375,
1225
+ "logps/rejected": -292.7288513183594,
1226
+ "loss": 0.4282,
1227
+ "rewards/accuracies": 0.824999988079071,
1228
+ "rewards/chosen": -1.087530493736267,
1229
+ "rewards/margins": 1.8494288921356201,
1230
+ "rewards/rejected": -2.9369590282440186,
1231
+ "step": 730
1232
+ },
1233
+ {
1234
+ "epoch": 1.29,
1235
+ "grad_norm": 7.472696962004615,
1236
+ "learning_rate": 1.6899546983993814e-07,
1237
+ "logits/chosen": -2.207268476486206,
1238
+ "logits/rejected": -2.2193145751953125,
1239
+ "logps/chosen": -333.41241455078125,
1240
+ "logps/rejected": -339.80670166015625,
1241
+ "loss": 0.4344,
1242
+ "rewards/accuracies": 0.78125,
1243
+ "rewards/chosen": -1.152045488357544,
1244
+ "rewards/margins": 1.8111851215362549,
1245
+ "rewards/rejected": -2.963230609893799,
1246
+ "step": 740
1247
+ },
1248
+ {
1249
+ "epoch": 1.31,
1250
+ "grad_norm": 7.854994688865555,
1251
+ "learning_rate": 1.618411361371193e-07,
1252
+ "logits/chosen": -2.241340398788452,
1253
+ "logits/rejected": -2.2305684089660645,
1254
+ "logps/chosen": -322.3190612792969,
1255
+ "logps/rejected": -343.23797607421875,
1256
+ "loss": 0.4631,
1257
+ "rewards/accuracies": 0.768750011920929,
1258
+ "rewards/chosen": -1.183625340461731,
1259
+ "rewards/margins": 1.5439379215240479,
1260
+ "rewards/rejected": -2.7275633811950684,
1261
+ "step": 750
1262
+ },
1263
+ {
1264
+ "epoch": 1.32,
1265
+ "grad_norm": 8.512012836575265,
1266
+ "learning_rate": 1.547683350946268e-07,
1267
+ "logits/chosen": -2.2200608253479004,
1268
+ "logits/rejected": -2.225961923599243,
1269
+ "logps/chosen": -326.2708740234375,
1270
+ "logps/rejected": -365.87689208984375,
1271
+ "loss": 0.4516,
1272
+ "rewards/accuracies": 0.887499988079071,
1273
+ "rewards/chosen": -1.0509939193725586,
1274
+ "rewards/margins": 2.1092488765716553,
1275
+ "rewards/rejected": -3.160243034362793,
1276
+ "step": 760
1277
+ },
1278
+ {
1279
+ "epoch": 1.34,
1280
+ "grad_norm": 7.243888365117908,
1281
+ "learning_rate": 1.477836079070687e-07,
1282
+ "logits/chosen": -2.1900453567504883,
1283
+ "logits/rejected": -2.194736957550049,
1284
+ "logps/chosen": -297.6527404785156,
1285
+ "logps/rejected": -324.55035400390625,
1286
+ "loss": 0.4447,
1287
+ "rewards/accuracies": 0.768750011920929,
1288
+ "rewards/chosen": -1.1286150217056274,
1289
+ "rewards/margins": 1.6178051233291626,
1290
+ "rewards/rejected": -2.74642014503479,
1291
+ "step": 770
1292
+ },
1293
+ {
1294
+ "epoch": 1.36,
1295
+ "grad_norm": 8.27216995738527,
1296
+ "learning_rate": 1.4089341431501228e-07,
1297
+ "logits/chosen": -2.1735267639160156,
1298
+ "logits/rejected": -2.2048020362854004,
1299
+ "logps/chosen": -341.09429931640625,
1300
+ "logps/rejected": -362.09521484375,
1301
+ "loss": 0.4533,
1302
+ "rewards/accuracies": 0.78125,
1303
+ "rewards/chosen": -1.1955785751342773,
1304
+ "rewards/margins": 2.0903077125549316,
1305
+ "rewards/rejected": -3.285886287689209,
1306
+ "step": 780
1307
+ },
1308
+ {
1309
+ "epoch": 1.38,
1310
+ "grad_norm": 6.73898031572128,
1311
+ "learning_rate": 1.3410412663077075e-07,
1312
+ "logits/chosen": -2.203127384185791,
1313
+ "logits/rejected": -2.1927337646484375,
1314
+ "logps/chosen": -297.426513671875,
1315
+ "logps/rejected": -331.25115966796875,
1316
+ "loss": 0.4387,
1317
+ "rewards/accuracies": 0.7749999761581421,
1318
+ "rewards/chosen": -1.2261813879013062,
1319
+ "rewards/margins": 1.633111596107483,
1320
+ "rewards/rejected": -2.859293222427368,
1321
+ "step": 790
1322
+ },
1323
+ {
1324
+ "epoch": 1.39,
1325
+ "grad_norm": 8.710779913480414,
1326
+ "learning_rate": 1.2742202384504757e-07,
1327
+ "logits/chosen": -2.2027111053466797,
1328
+ "logits/rejected": -2.2479450702667236,
1329
+ "logps/chosen": -331.02740478515625,
1330
+ "logps/rejected": -364.1605224609375,
1331
+ "loss": 0.471,
1332
+ "rewards/accuracies": 0.768750011920929,
1333
+ "rewards/chosen": -1.1860400438308716,
1334
+ "rewards/margins": 1.6458642482757568,
1335
+ "rewards/rejected": -2.831904172897339,
1336
+ "step": 800
1337
+ },
1338
+ {
1339
+ "epoch": 1.39,
1340
+ "eval_logits/chosen": -2.299058437347412,
1341
+ "eval_logits/rejected": -2.3809187412261963,
1342
+ "eval_logps/chosen": -316.5019226074219,
1343
+ "eval_logps/rejected": -321.5930480957031,
1344
+ "eval_loss": 0.4368092715740204,
1345
+ "eval_rewards/accuracies": 0.7689873576164246,
1346
+ "eval_rewards/chosen": -0.9616590142250061,
1347
+ "eval_rewards/margins": 1.7827986478805542,
1348
+ "eval_rewards/rejected": -2.744457483291626,
1349
+ "eval_runtime": 117.6971,
1350
+ "eval_samples_per_second": 21.241,
1351
+ "eval_steps_per_second": 0.671,
1352
+ "step": 800
1353
+ },
1354
+ {
1355
+ "epoch": 1.41,
1356
+ "grad_norm": 7.470099061565703,
1357
+ "learning_rate": 1.208532858198875e-07,
1358
+ "logits/chosen": -2.19744610786438,
1359
+ "logits/rejected": -2.2142927646636963,
1360
+ "logps/chosen": -326.9881591796875,
1361
+ "logps/rejected": -365.5570983886719,
1362
+ "loss": 0.4298,
1363
+ "rewards/accuracies": 0.78125,
1364
+ "rewards/chosen": -1.2280399799346924,
1365
+ "rewards/margins": 1.8768724203109741,
1366
+ "rewards/rejected": -3.104912519454956,
1367
+ "step": 810
1368
+ },
1369
+ {
1370
+ "epoch": 1.43,
1371
+ "grad_norm": 6.878834958249367,
1372
+ "learning_rate": 1.1440398757330589e-07,
1373
+ "logits/chosen": -2.16925048828125,
1374
+ "logits/rejected": -2.2376468181610107,
1375
+ "logps/chosen": -333.90338134765625,
1376
+ "logps/rejected": -332.63299560546875,
1377
+ "loss": 0.4255,
1378
+ "rewards/accuracies": 0.84375,
1379
+ "rewards/chosen": -1.1526517868041992,
1380
+ "rewards/margins": 1.8892349004745483,
1381
+ "rewards/rejected": -3.041886806488037,
1382
+ "step": 820
1383
+ },
1384
+ {
1385
+ "epoch": 1.45,
1386
+ "grad_norm": 6.741909833219955,
1387
+ "learning_rate": 1.0808009366088155e-07,
1388
+ "logits/chosen": -2.205474853515625,
1389
+ "logits/rejected": -2.2162137031555176,
1390
+ "logps/chosen": -325.2530212402344,
1391
+ "logps/rejected": -318.7095031738281,
1392
+ "loss": 0.4789,
1393
+ "rewards/accuracies": 0.762499988079071,
1394
+ "rewards/chosen": -1.2332874536514282,
1395
+ "rewards/margins": 1.781602144241333,
1396
+ "rewards/rejected": -3.0148894786834717,
1397
+ "step": 830
1398
+ },
1399
+ {
1400
+ "epoch": 1.46,
1401
+ "grad_norm": 7.3191732297036225,
1402
+ "learning_rate": 1.0188745265950985e-07,
1403
+ "logits/chosen": -2.229236602783203,
1404
+ "logits/rejected": -2.2648098468780518,
1405
+ "logps/chosen": -308.75396728515625,
1406
+ "logps/rejected": -331.4061279296875,
1407
+ "loss": 0.4429,
1408
+ "rewards/accuracies": 0.78125,
1409
+ "rewards/chosen": -1.206955075263977,
1410
+ "rewards/margins": 1.5560095310211182,
1411
+ "rewards/rejected": -2.7629647254943848,
1412
+ "step": 840
1413
+ },
1414
+ {
1415
+ "epoch": 1.48,
1416
+ "grad_norm": 8.26769354330985,
1417
+ "learning_rate": 9.583179175841666e-08,
1418
+ "logits/chosen": -2.199425220489502,
1419
+ "logits/rejected": -2.1700549125671387,
1420
+ "logps/chosen": -314.6337890625,
1421
+ "logps/rejected": -338.1066589355469,
1422
+ "loss": 0.4538,
1423
+ "rewards/accuracies": 0.8187500238418579,
1424
+ "rewards/chosen": -1.110811710357666,
1425
+ "rewards/margins": 1.5624140501022339,
1426
+ "rewards/rejected": -2.6732258796691895,
1427
+ "step": 850
1428
+ },
1429
+ {
1430
+ "epoch": 1.5,
1431
+ "grad_norm": 7.000649639637843,
1432
+ "learning_rate": 8.991871146243696e-08,
1433
+ "logits/chosen": -2.217109203338623,
1434
+ "logits/rejected": -2.2137506008148193,
1435
+ "logps/chosen": -289.3517761230469,
1436
+ "logps/rejected": -343.0474853515625,
1437
+ "loss": 0.4358,
1438
+ "rewards/accuracies": 0.768750011920929,
1439
+ "rewards/chosen": -1.2585060596466064,
1440
+ "rewards/margins": 1.6992295980453491,
1441
+ "rewards/rejected": -2.957735538482666,
1442
+ "step": 860
1443
+ },
1444
+ {
1445
+ "epoch": 1.52,
1446
+ "grad_norm": 6.509621971875314,
1447
+ "learning_rate": 8.415368041245513e-08,
1448
+ "logits/chosen": -2.205955743789673,
1449
+ "logits/rejected": -2.2050979137420654,
1450
+ "logps/chosen": -284.9857482910156,
1451
+ "logps/rejected": -335.174560546875,
1452
+ "loss": 0.439,
1453
+ "rewards/accuracies": 0.78125,
1454
+ "rewards/chosen": -1.2468931674957275,
1455
+ "rewards/margins": 1.6571691036224365,
1456
+ "rewards/rejected": -2.904062032699585,
1457
+ "step": 870
1458
+ },
1459
+ {
1460
+ "epoch": 1.53,
1461
+ "grad_norm": 12.101058956711388,
1462
+ "learning_rate": 7.854203032779894e-08,
1463
+ "logits/chosen": -2.2182745933532715,
1464
+ "logits/rejected": -2.268409490585327,
1465
+ "logps/chosen": -342.60528564453125,
1466
+ "logps/rejected": -349.5751953125,
1467
+ "loss": 0.4508,
1468
+ "rewards/accuracies": 0.75,
1469
+ "rewards/chosen": -1.2632758617401123,
1470
+ "rewards/margins": 1.6573642492294312,
1471
+ "rewards/rejected": -2.920640468597412,
1472
+ "step": 880
1473
+ },
1474
+ {
1475
+ "epoch": 1.55,
1476
+ "grad_norm": 7.466182101903001,
1477
+ "learning_rate": 7.308895107526317e-08,
1478
+ "logits/chosen": -2.1878528594970703,
1479
+ "logits/rejected": -2.2148966789245605,
1480
+ "logps/chosen": -329.15374755859375,
1481
+ "logps/rejected": -361.855224609375,
1482
+ "loss": 0.4403,
1483
+ "rewards/accuracies": 0.7749999761581421,
1484
+ "rewards/chosen": -1.0873197317123413,
1485
+ "rewards/margins": 1.8896510601043701,
1486
+ "rewards/rejected": -2.976970672607422,
1487
+ "step": 890
1488
+ },
1489
+ {
1490
+ "epoch": 1.57,
1491
+ "grad_norm": 7.096542397688041,
1492
+ "learning_rate": 6.779948586932389e-08,
1493
+ "logits/chosen": -2.1874141693115234,
1494
+ "logits/rejected": -2.2552855014801025,
1495
+ "logps/chosen": -339.11761474609375,
1496
+ "logps/rejected": -330.0443115234375,
1497
+ "loss": 0.4485,
1498
+ "rewards/accuracies": 0.7437499761581421,
1499
+ "rewards/chosen": -1.2248756885528564,
1500
+ "rewards/margins": 1.6217113733291626,
1501
+ "rewards/rejected": -2.8465871810913086,
1502
+ "step": 900
1503
+ },
1504
+ {
1505
+ "epoch": 1.57,
1506
+ "eval_logits/chosen": -2.3004236221313477,
1507
+ "eval_logits/rejected": -2.3815455436706543,
1508
+ "eval_logps/chosen": -316.2496643066406,
1509
+ "eval_logps/rejected": -321.8916015625,
1510
+ "eval_loss": 0.435116708278656,
1511
+ "eval_rewards/accuracies": 0.7721518874168396,
1512
+ "eval_rewards/chosen": -0.9490465521812439,
1513
+ "eval_rewards/margins": 1.8103375434875488,
1514
+ "eval_rewards/rejected": -2.7593843936920166,
1515
+ "eval_runtime": 116.8837,
1516
+ "eval_samples_per_second": 21.389,
1517
+ "eval_steps_per_second": 0.676,
1518
+ "step": 900
1519
+ },
1520
+ {
1521
+ "epoch": 1.58,
1522
+ "grad_norm": 8.40277649092231,
1523
+ "learning_rate": 6.267852660798256e-08,
1524
+ "logits/chosen": -2.243600845336914,
1525
+ "logits/rejected": -2.1874167919158936,
1526
+ "logps/chosen": -264.4030456542969,
1527
+ "logps/rejected": -318.32659912109375,
1528
+ "loss": 0.4608,
1529
+ "rewards/accuracies": 0.78125,
1530
+ "rewards/chosen": -1.2417147159576416,
1531
+ "rewards/margins": 1.8263641595840454,
1532
+ "rewards/rejected": -3.0680789947509766,
1533
+ "step": 910
1534
+ },
1535
+ {
1536
+ "epoch": 1.6,
1537
+ "grad_norm": 7.94424609527543,
1538
+ "learning_rate": 5.7730809348553315e-08,
1539
+ "logits/chosen": -2.226975679397583,
1540
+ "logits/rejected": -2.2741405963897705,
1541
+ "logps/chosen": -309.5135803222656,
1542
+ "logps/rejected": -325.623779296875,
1543
+ "loss": 0.4366,
1544
+ "rewards/accuracies": 0.8187500238418579,
1545
+ "rewards/chosen": -1.201520562171936,
1546
+ "rewards/margins": 1.7903496026992798,
1547
+ "rewards/rejected": -2.991870164871216,
1548
+ "step": 920
1549
+ },
1550
+ {
1551
+ "epoch": 1.62,
1552
+ "grad_norm": 7.128082903747095,
1553
+ "learning_rate": 5.296090992757746e-08,
1554
+ "logits/chosen": -2.206270217895508,
1555
+ "logits/rejected": -2.201709747314453,
1556
+ "logps/chosen": -315.83282470703125,
1557
+ "logps/rejected": -364.1822814941406,
1558
+ "loss": 0.443,
1559
+ "rewards/accuracies": 0.78125,
1560
+ "rewards/chosen": -1.2698633670806885,
1561
+ "rewards/margins": 1.9889068603515625,
1562
+ "rewards/rejected": -3.258769989013672,
1563
+ "step": 930
1564
+ },
1565
+ {
1566
+ "epoch": 1.64,
1567
+ "grad_norm": 7.514968298076625,
1568
+ "learning_rate": 4.8373239728916326e-08,
1569
+ "logits/chosen": -2.1958649158477783,
1570
+ "logits/rejected": -2.1885480880737305,
1571
+ "logps/chosen": -288.6278381347656,
1572
+ "logps/rejected": -353.77984619140625,
1573
+ "loss": 0.4283,
1574
+ "rewards/accuracies": 0.8374999761581421,
1575
+ "rewards/chosen": -1.0743087530136108,
1576
+ "rewards/margins": 2.0363688468933105,
1577
+ "rewards/rejected": -3.1106772422790527,
1578
+ "step": 940
1579
+ },
1580
+ {
1581
+ "epoch": 1.65,
1582
+ "grad_norm": 11.932999852584523,
1583
+ "learning_rate": 4.397204160393628e-08,
1584
+ "logits/chosen": -2.192418336868286,
1585
+ "logits/rejected": -2.225080966949463,
1586
+ "logps/chosen": -281.90093994140625,
1587
+ "logps/rejected": -319.2328186035156,
1588
+ "loss": 0.4456,
1589
+ "rewards/accuracies": 0.78125,
1590
+ "rewards/chosen": -1.106330156326294,
1591
+ "rewards/margins": 1.9500024318695068,
1592
+ "rewards/rejected": -3.056332588195801,
1593
+ "step": 950
1594
+ },
1595
+ {
1596
+ "epoch": 1.67,
1597
+ "grad_norm": 7.8327307820265375,
1598
+ "learning_rate": 3.9761385947558845e-08,
1599
+ "logits/chosen": -2.2083628177642822,
1600
+ "logits/rejected": -2.212167978286743,
1601
+ "logps/chosen": -305.710693359375,
1602
+ "logps/rejected": -331.8744201660156,
1603
+ "loss": 0.4156,
1604
+ "rewards/accuracies": 0.8125,
1605
+ "rewards/chosen": -1.0697580575942993,
1606
+ "rewards/margins": 2.013720989227295,
1607
+ "rewards/rejected": -3.083479404449463,
1608
+ "step": 960
1609
+ },
1610
+ {
1611
+ "epoch": 1.69,
1612
+ "grad_norm": 8.141593893840476,
1613
+ "learning_rate": 3.574516693380511e-08,
1614
+ "logits/chosen": -2.177604913711548,
1615
+ "logits/rejected": -2.187506914138794,
1616
+ "logps/chosen": -293.26287841796875,
1617
+ "logps/rejected": -328.9276123046875,
1618
+ "loss": 0.4442,
1619
+ "rewards/accuracies": 0.7749999761581421,
1620
+ "rewards/chosen": -1.204113245010376,
1621
+ "rewards/margins": 1.9541772603988647,
1622
+ "rewards/rejected": -3.158290386199951,
1623
+ "step": 970
1624
+ },
1625
+ {
1626
+ "epoch": 1.71,
1627
+ "grad_norm": 8.19223484059336,
1628
+ "learning_rate": 3.192709891431586e-08,
1629
+ "logits/chosen": -2.1618940830230713,
1630
+ "logits/rejected": -2.1518969535827637,
1631
+ "logps/chosen": -324.14654541015625,
1632
+ "logps/rejected": -348.9007263183594,
1633
+ "loss": 0.4289,
1634
+ "rewards/accuracies": 0.762499988079071,
1635
+ "rewards/chosen": -1.1970574855804443,
1636
+ "rewards/margins": 1.7036092281341553,
1637
+ "rewards/rejected": -2.9006664752960205,
1638
+ "step": 980
1639
+ },
1640
+ {
1641
+ "epoch": 1.72,
1642
+ "grad_norm": 7.729663351327424,
1643
+ "learning_rate": 2.8310712983178524e-08,
1644
+ "logits/chosen": -2.204336404800415,
1645
+ "logits/rejected": -2.2102417945861816,
1646
+ "logps/chosen": -307.18817138671875,
1647
+ "logps/rejected": -341.29840087890625,
1648
+ "loss": 0.4431,
1649
+ "rewards/accuracies": 0.731249988079071,
1650
+ "rewards/chosen": -1.2059943675994873,
1651
+ "rewards/margins": 1.5643813610076904,
1652
+ "rewards/rejected": -2.7703757286071777,
1653
+ "step": 990
1654
+ },
1655
+ {
1656
+ "epoch": 1.74,
1657
+ "grad_norm": 8.086893591573848,
1658
+ "learning_rate": 2.4899353711237247e-08,
1659
+ "logits/chosen": -2.1703457832336426,
1660
+ "logits/rejected": -2.2025086879730225,
1661
+ "logps/chosen": -313.03167724609375,
1662
+ "logps/rejected": -340.02264404296875,
1663
+ "loss": 0.4411,
1664
+ "rewards/accuracies": 0.8125,
1665
+ "rewards/chosen": -1.1995328664779663,
1666
+ "rewards/margins": 1.7753593921661377,
1667
+ "rewards/rejected": -2.9748923778533936,
1668
+ "step": 1000
1669
+ },
1670
+ {
1671
+ "epoch": 1.74,
1672
+ "eval_logits/chosen": -2.3010759353637695,
1673
+ "eval_logits/rejected": -2.3822624683380127,
1674
+ "eval_logps/chosen": -315.85467529296875,
1675
+ "eval_logps/rejected": -321.64093017578125,
1676
+ "eval_loss": 0.4348324239253998,
1677
+ "eval_rewards/accuracies": 0.7658227682113647,
1678
+ "eval_rewards/chosen": -0.9292957782745361,
1679
+ "eval_rewards/margins": 1.8175575733184814,
1680
+ "eval_rewards/rejected": -2.7468531131744385,
1681
+ "eval_runtime": 116.7271,
1682
+ "eval_samples_per_second": 21.417,
1683
+ "eval_steps_per_second": 0.677,
1684
+ "step": 1000
1685
+ },
1686
+ {
1687
+ "epoch": 1.76,
1688
+ "grad_norm": 6.876861850417737,
1689
+ "learning_rate": 2.1696176052907105e-08,
1690
+ "logits/chosen": -2.2324881553649902,
1691
+ "logits/rejected": -2.219320774078369,
1692
+ "logps/chosen": -313.92529296875,
1693
+ "logps/rejected": -350.4786682128906,
1694
+ "loss": 0.4293,
1695
+ "rewards/accuracies": 0.7749999761581421,
1696
+ "rewards/chosen": -1.2920801639556885,
1697
+ "rewards/margins": 1.7782132625579834,
1698
+ "rewards/rejected": -3.070293426513672,
1699
+ "step": 1010
1700
+ },
1701
+ {
1702
+ "epoch": 1.78,
1703
+ "grad_norm": 7.677036235689073,
1704
+ "learning_rate": 1.8704142428352528e-08,
1705
+ "logits/chosen": -2.190504789352417,
1706
+ "logits/rejected": -2.2708182334899902,
1707
+ "logps/chosen": -332.3207702636719,
1708
+ "logps/rejected": -330.7673034667969,
1709
+ "loss": 0.4507,
1710
+ "rewards/accuracies": 0.78125,
1711
+ "rewards/chosen": -1.1538569927215576,
1712
+ "rewards/margins": 1.831903100013733,
1713
+ "rewards/rejected": -2.98576021194458,
1714
+ "step": 1020
1715
+ },
1716
+ {
1717
+ "epoch": 1.79,
1718
+ "grad_norm": 6.9135152123839365,
1719
+ "learning_rate": 1.592601998372886e-08,
1720
+ "logits/chosen": -2.195899248123169,
1721
+ "logits/rejected": -2.1886417865753174,
1722
+ "logps/chosen": -296.1393127441406,
1723
+ "logps/rejected": -334.1782531738281,
1724
+ "loss": 0.4433,
1725
+ "rewards/accuracies": 0.7749999761581421,
1726
+ "rewards/chosen": -1.1621323823928833,
1727
+ "rewards/margins": 1.852728247642517,
1728
+ "rewards/rejected": -3.0148606300354004,
1729
+ "step": 1030
1730
+ },
1731
+ {
1732
+ "epoch": 1.81,
1733
+ "grad_norm": 7.569188340166907,
1734
+ "learning_rate": 1.336437803202059e-08,
1735
+ "logits/chosen": -2.228149175643921,
1736
+ "logits/rejected": -2.2481937408447266,
1737
+ "logps/chosen": -310.6243896484375,
1738
+ "logps/rejected": -328.8221130371094,
1739
+ "loss": 0.4301,
1740
+ "rewards/accuracies": 0.8187500238418579,
1741
+ "rewards/chosen": -1.1201293468475342,
1742
+ "rewards/margins": 1.740098237991333,
1743
+ "rewards/rejected": -2.860227584838867,
1744
+ "step": 1040
1745
+ },
1746
+ {
1747
+ "epoch": 1.83,
1748
+ "grad_norm": 8.234591785864946,
1749
+ "learning_rate": 1.102158567684336e-08,
1750
+ "logits/chosen": -2.192157745361328,
1751
+ "logits/rejected": -2.194495677947998,
1752
+ "logps/chosen": -302.17559814453125,
1753
+ "logps/rejected": -339.31085205078125,
1754
+ "loss": 0.4208,
1755
+ "rewards/accuracies": 0.800000011920929,
1756
+ "rewards/chosen": -1.169334053993225,
1757
+ "rewards/margins": 1.9188495874404907,
1758
+ "rewards/rejected": -3.088183879852295,
1759
+ "step": 1050
1760
+ },
1761
+ {
1762
+ "epoch": 1.85,
1763
+ "grad_norm": 7.558156377001319,
1764
+ "learning_rate": 8.899809621407045e-09,
1765
+ "logits/chosen": -2.1941123008728027,
1766
+ "logits/rejected": -2.2328617572784424,
1767
+ "logps/chosen": -275.9726867675781,
1768
+ "logps/rejected": -308.49212646484375,
1769
+ "loss": 0.4256,
1770
+ "rewards/accuracies": 0.7437499761581421,
1771
+ "rewards/chosen": -1.2298341989517212,
1772
+ "rewards/margins": 1.8165092468261719,
1773
+ "rewards/rejected": -3.0463433265686035,
1774
+ "step": 1060
1775
+ },
1776
+ {
1777
+ "epoch": 1.86,
1778
+ "grad_norm": 7.958869766486015,
1779
+ "learning_rate": 7.001012164666392e-09,
1780
+ "logits/chosen": -2.181845188140869,
1781
+ "logits/rejected": -2.1818060874938965,
1782
+ "logps/chosen": -332.1898498535156,
1783
+ "logps/rejected": -357.20062255859375,
1784
+ "loss": 0.4367,
1785
+ "rewards/accuracies": 0.7437499761581421,
1786
+ "rewards/chosen": -1.14080810546875,
1787
+ "rewards/margins": 1.8624111413955688,
1788
+ "rewards/rejected": -3.0032193660736084,
1789
+ "step": 1070
1790
+ },
1791
+ {
1792
+ "epoch": 1.88,
1793
+ "grad_norm": 7.273641922997009,
1794
+ "learning_rate": 5.326949386512764e-09,
1795
+ "logits/chosen": -2.1961543560028076,
1796
+ "logits/rejected": -2.1588053703308105,
1797
+ "logps/chosen": -314.35601806640625,
1798
+ "logps/rejected": -367.77716064453125,
1799
+ "loss": 0.4241,
1800
+ "rewards/accuracies": 0.7749999761581421,
1801
+ "rewards/chosen": -1.0949821472167969,
1802
+ "rewards/margins": 2.1694438457489014,
1803
+ "rewards/rejected": -3.2644259929656982,
1804
+ "step": 1080
1805
+ },
1806
+ {
1807
+ "epoch": 1.9,
1808
+ "grad_norm": 7.683716889020659,
1809
+ "learning_rate": 3.879169523684639e-09,
1810
+ "logits/chosen": -2.182697296142578,
1811
+ "logits/rejected": -2.188960313796997,
1812
+ "logps/chosen": -294.359375,
1813
+ "logps/rejected": -334.6479187011719,
1814
+ "loss": 0.4398,
1815
+ "rewards/accuracies": 0.768750011920929,
1816
+ "rewards/chosen": -1.1956040859222412,
1817
+ "rewards/margins": 1.873143196105957,
1818
+ "rewards/rejected": -3.0687472820281982,
1819
+ "step": 1090
1820
+ },
1821
+ {
1822
+ "epoch": 1.92,
1823
+ "grad_norm": 8.782560420891002,
1824
+ "learning_rate": 2.65901153789963e-09,
1825
+ "logits/chosen": -2.2032766342163086,
1826
+ "logits/rejected": -2.225320339202881,
1827
+ "logps/chosen": -302.3791198730469,
1828
+ "logps/rejected": -329.71710205078125,
1829
+ "loss": 0.4499,
1830
+ "rewards/accuracies": 0.793749988079071,
1831
+ "rewards/chosen": -1.1408864259719849,
1832
+ "rewards/margins": 1.8591581583023071,
1833
+ "rewards/rejected": -3.000044345855713,
1834
+ "step": 1100
1835
+ },
1836
+ {
1837
+ "epoch": 1.92,
1838
+ "eval_logits/chosen": -2.301210641860962,
1839
+ "eval_logits/rejected": -2.382761240005493,
1840
+ "eval_logps/chosen": -316.2319641113281,
1841
+ "eval_logps/rejected": -322.2369079589844,
1842
+ "eval_loss": 0.4347890019416809,
1843
+ "eval_rewards/accuracies": 0.7658227682113647,
1844
+ "eval_rewards/chosen": -0.9481591582298279,
1845
+ "eval_rewards/margins": 1.8284918069839478,
1846
+ "eval_rewards/rejected": -2.776650905609131,
1847
+ "eval_runtime": 116.8141,
1848
+ "eval_samples_per_second": 21.402,
1849
+ "eval_steps_per_second": 0.676,
1850
+ "step": 1100
1851
+ },
1852
+ {
1853
+ "epoch": 1.93,
1854
+ "grad_norm": 7.9306460884331775,
1855
+ "learning_rate": 1.6676038775320089e-09,
1856
+ "logits/chosen": -2.1896042823791504,
1857
+ "logits/rejected": -2.1769232749938965,
1858
+ "logps/chosen": -319.85491943359375,
1859
+ "logps/rejected": -358.85296630859375,
1860
+ "loss": 0.4625,
1861
+ "rewards/accuracies": 0.7875000238418579,
1862
+ "rewards/chosen": -1.1989891529083252,
1863
+ "rewards/margins": 1.7593486309051514,
1864
+ "rewards/rejected": -2.9583375453948975,
1865
+ "step": 1110
1866
+ },
1867
+ {
1868
+ "epoch": 1.95,
1869
+ "grad_norm": 6.9870920606828735,
1870
+ "learning_rate": 9.058634339806914e-10,
1871
+ "logits/chosen": -2.190636396408081,
1872
+ "logits/rejected": -2.2027156352996826,
1873
+ "logps/chosen": -313.0627746582031,
1874
+ "logps/rejected": -347.0715026855469,
1875
+ "loss": 0.4495,
1876
+ "rewards/accuracies": 0.768750011920929,
1877
+ "rewards/chosen": -1.2059122323989868,
1878
+ "rewards/margins": 1.832135558128357,
1879
+ "rewards/rejected": -3.0380477905273438,
1880
+ "step": 1120
1881
+ },
1882
+ {
1883
+ "epoch": 1.97,
1884
+ "grad_norm": 8.257014654020075,
1885
+ "learning_rate": 3.74494693693439e-10,
1886
+ "logits/chosen": -2.2022292613983154,
1887
+ "logits/rejected": -2.2187328338623047,
1888
+ "logps/chosen": -282.579345703125,
1889
+ "logps/rejected": -332.579345703125,
1890
+ "loss": 0.4331,
1891
+ "rewards/accuracies": 0.7562500238418579,
1892
+ "rewards/chosen": -1.009050965309143,
1893
+ "rewards/margins": 1.674533486366272,
1894
+ "rewards/rejected": -2.683584690093994,
1895
+ "step": 1130
1896
+ },
1897
+ {
1898
+ "epoch": 1.99,
1899
+ "grad_norm": 8.586752011175628,
1900
+ "learning_rate": 7.39890866310322e-11,
1901
+ "logits/chosen": -2.1978275775909424,
1902
+ "logits/rejected": -2.208787441253662,
1903
+ "logps/chosen": -313.6376953125,
1904
+ "logps/rejected": -326.4861145019531,
1905
+ "loss": 0.4289,
1906
+ "rewards/accuracies": 0.6875,
1907
+ "rewards/chosen": -1.2633991241455078,
1908
+ "rewards/margins": 1.3990856409072876,
1909
+ "rewards/rejected": -2.662484645843506,
1910
+ "step": 1140
1911
+ },
1912
+ {
1913
+ "epoch": 2.0,
1914
+ "step": 1148,
1915
+ "total_flos": 0.0,
1916
+ "train_loss": 0.5184940074794384,
1917
+ "train_runtime": 19743.6623,
1918
+ "train_samples_per_second": 7.445,
1919
+ "train_steps_per_second": 0.058
1920
+ }
1921
+ ],
1922
+ "logging_steps": 10,
1923
+ "max_steps": 1148,
1924
+ "num_input_tokens_seen": 0,
1925
+ "num_train_epochs": 2,
1926
+ "save_steps": 100,
1927
+ "total_flos": 0.0,
1928
+ "train_batch_size": 2,
1929
+ "trial_name": null,
1930
+ "trial_params": null
1931
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff