Hugging Face's logo

nicolof88
/
mistral7b-spider-ft

Text Generation
TensorBoard
Safetensors
Trained with AutoTrain
conversational
Inference Endpoints
Model card Files Metrics Community
nicolof88 commited on
Commit
32de623
•
1 Parent(s): da9c083

Upload folder using huggingface_hub

Browse files
Files changed (25) hide show
  1. README.md +40 -0
  2. adapter_config.json +33 -0
  3. adapter_model.safetensors +3 -0
  4. checkpoint-9676/README.md +202 -0
  5. checkpoint-9676/adapter_config.json +33 -0
  6. checkpoint-9676/adapter_model.safetensors +3 -0
  7. checkpoint-9676/optimizer.pt +3 -0
  8. checkpoint-9676/pytorch_model.bin +3 -0
  9. checkpoint-9676/rng_state.pth +3 -0
  10. checkpoint-9676/scheduler.pt +3 -0
  11. checkpoint-9676/special_tokens_map.json +24 -0
  12. checkpoint-9676/tokenizer.json +0 -0
  13. checkpoint-9676/tokenizer.model +3 -0
  14. checkpoint-9676/tokenizer_config.json +43 -0
  15. checkpoint-9676/trainer_state.json +2730 -0
  16. checkpoint-9676/training_args.bin +3 -0
  17. handler.py +32 -0
  18. requirements.txt +2 -0
  19. runs/Mar21_02-37-23_r-nicolof88-mistral7b-spider-at-188n2lyh-a2f2a-vnlru/events.out.tfevents.1710988652.r-nicolof88-mistral7b-spider-at-188n2lyh-a2f2a-vnlru.98.0 +2 -2
  20. special_tokens_map.json +24 -0
  21. tokenizer.json +0 -0
  22. tokenizer.model +3 -0
  23. tokenizer_config.json +43 -0
  24. training_args.bin +3 -0
  25. training_params.json +47 -0
README.md ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - autotrain
4
+ - text-generation
5
+ widget:
6
+ - text: "I love AutoTrain because "
7
+ license: other
8
+ ---
9
+
10
+ # Model Trained Using AutoTrain
11
+
12
+ This model was trained using AutoTrain. For more information, please visit [AutoTrain](https://hf.co/docs/autotrain).
13
+
14
+ # Usage
15
+
16
+ ```python
17
+
18
+ from transformers import AutoModelForCausalLM, AutoTokenizer
19
+
20
+ model_path = "PATH_TO_THIS_REPO"
21
+
22
+ tokenizer = AutoTokenizer.from_pretrained(model_path)
23
+ model = AutoModelForCausalLM.from_pretrained(
24
+ model_path,
25
+ device_map="auto",
26
+ torch_dtype='auto'
27
+ ).eval()
28
+
29
+ # Prompt content: "hi"
30
+ messages = [
31
+ {"role": "user", "content": "hi"}
32
+ ]
33
+
34
+ input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
35
+ output_ids = model.generate(input_ids.to('cuda'))
36
+ response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
37
+
38
+ # Model response: "Hello! How can I assist you today?"
39
+ print(response)
40
+ ```
adapter_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "mistralai/Mistral-7B-Instruct-v0.2",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "loftq_config": {},
12
+ "lora_alpha": 16,
13
+ "lora_dropout": 0.05,
14
+ "megatron_config": null,
15
+ "megatron_core": "megatron.core",
16
+ "modules_to_save": null,
17
+ "peft_type": "LORA",
18
+ "r": 64,
19
+ "rank_pattern": {},
20
+ "revision": null,
21
+ "target_modules": [
22
+ "q_proj",
23
+ "down_proj",
24
+ "v_proj",
25
+ "up_proj",
26
+ "k_proj",
27
+ "gate_proj",
28
+ "o_proj"
29
+ ],
30
+ "task_type": "CAUSAL_LM",
31
+ "use_dora": false,
32
+ "use_rslora": false
33
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf35931b3c5aed981d5e1d9a71d0340a4761c5157b8fccf28d16da2b1c080439
3
+ size 671149168
checkpoint-9676/README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ base_model: mistralai/Mistral-7B-Instruct-v0.2
4
+ ---
5
+
6
+ # Model Card for Model ID
7
+
8
+ <!-- Provide a quick summary of what the model is/does. -->
9
+
10
+
11
+
12
+ ## Model Details
13
+
14
+ ### Model Description
15
+
16
+ <!-- Provide a longer summary of what this model is. -->
17
+
18
+
19
+
20
+ - **Developed by:** [More Information Needed]
21
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
200
+ ### Framework versions
201
+
202
+ - PEFT 0.9.0
checkpoint-9676/adapter_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "mistralai/Mistral-7B-Instruct-v0.2",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "loftq_config": {},
12
+ "lora_alpha": 16,
13
+ "lora_dropout": 0.05,
14
+ "megatron_config": null,
15
+ "megatron_core": "megatron.core",
16
+ "modules_to_save": null,
17
+ "peft_type": "LORA",
18
+ "r": 64,
19
+ "rank_pattern": {},
20
+ "revision": null,
21
+ "target_modules": [
22
+ "q_proj",
23
+ "down_proj",
24
+ "v_proj",
25
+ "up_proj",
26
+ "k_proj",
27
+ "gate_proj",
28
+ "o_proj"
29
+ ],
30
+ "task_type": "CAUSAL_LM",
31
+ "use_dora": false,
32
+ "use_rslora": false
33
+ }
checkpoint-9676/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf35931b3c5aed981d5e1d9a71d0340a4761c5157b8fccf28d16da2b1c080439
3
+ size 671149168
checkpoint-9676/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc5176933f9170ac12c5c8c7757752889f45b015e76eac64a6dcc19cfbd31b82
3
+ size 1342555602
checkpoint-9676/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:049c26b844b79121ddd8379f7f69194e63f6fbf6aa007eeac0c66f17eebb8893
3
+ size 888
checkpoint-9676/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e3919a6c0eded6d6d7870145bb89f27445b92aae1ef7d2a38f1223b4e7820cfc
3
+ size 14244
checkpoint-9676/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60ef1f8511acc4ba8b3aa4c57afaff33980f2570f13a13ce4e7d272a72264763
3
+ size 1064
checkpoint-9676/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
checkpoint-9676/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-9676/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
checkpoint-9676/tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "additional_special_tokens": [],
31
+ "bos_token": "<s>",
32
+ "chat_template": "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ message['content'] + eos_token}}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}",
33
+ "clean_up_tokenization_spaces": false,
34
+ "eos_token": "</s>",
35
+ "legacy": true,
36
+ "model_max_length": 2048,
37
+ "pad_token": "</s>",
38
+ "sp_model_kwargs": {},
39
+ "spaces_between_special_tokens": false,
40
+ "tokenizer_class": "LlamaTokenizer",
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": false
43
+ }
checkpoint-9676/trainer_state.json ADDED
@@ -0,0 +1,2730 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 4.0,
5
+ "eval_steps": 500,
6
+ "global_step": 9676,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "grad_norm": 1.180048942565918,
14
+ "learning_rate": 7.747933884297521e-07,
15
+ "loss": 1.2211,
16
+ "step": 25
17
+ },
18
+ {
19
+ "epoch": 0.02,
20
+ "grad_norm": 0.8116350769996643,
21
+ "learning_rate": 1.5495867768595043e-06,
22
+ "loss": 1.1783,
23
+ "step": 50
24
+ },
25
+ {
26
+ "epoch": 0.03,
27
+ "grad_norm": 0.8945431709289551,
28
+ "learning_rate": 2.3243801652892563e-06,
29
+ "loss": 1.0768,
30
+ "step": 75
31
+ },
32
+ {
33
+ "epoch": 0.04,
34
+ "grad_norm": 0.4761434495449066,
35
+ "learning_rate": 3.0991735537190086e-06,
36
+ "loss": 0.8783,
37
+ "step": 100
38
+ },
39
+ {
40
+ "epoch": 0.05,
41
+ "grad_norm": 0.37374910712242126,
42
+ "learning_rate": 3.87396694214876e-06,
43
+ "loss": 0.8531,
44
+ "step": 125
45
+ },
46
+ {
47
+ "epoch": 0.06,
48
+ "grad_norm": 0.5040513277053833,
49
+ "learning_rate": 4.648760330578513e-06,
50
+ "loss": 0.7562,
51
+ "step": 150
52
+ },
53
+ {
54
+ "epoch": 0.07,
55
+ "grad_norm": 0.3926999270915985,
56
+ "learning_rate": 5.423553719008265e-06,
57
+ "loss": 0.6665,
58
+ "step": 175
59
+ },
60
+ {
61
+ "epoch": 0.08,
62
+ "grad_norm": 0.38587668538093567,
63
+ "learning_rate": 6.198347107438017e-06,
64
+ "loss": 0.6037,
65
+ "step": 200
66
+ },
67
+ {
68
+ "epoch": 0.09,
69
+ "grad_norm": 0.3665236830711365,
70
+ "learning_rate": 6.9731404958677686e-06,
71
+ "loss": 0.5731,
72
+ "step": 225
73
+ },
74
+ {
75
+ "epoch": 0.1,
76
+ "grad_norm": 0.43044623732566833,
77
+ "learning_rate": 7.74793388429752e-06,
78
+ "loss": 0.536,
79
+ "step": 250
80
+ },
81
+ {
82
+ "epoch": 0.11,
83
+ "grad_norm": 0.3693198561668396,
84
+ "learning_rate": 8.522727272727273e-06,
85
+ "loss": 0.5333,
86
+ "step": 275
87
+ },
88
+ {
89
+ "epoch": 0.12,
90
+ "grad_norm": 0.46105554699897766,
91
+ "learning_rate": 9.297520661157025e-06,
92
+ "loss": 0.4943,
93
+ "step": 300
94
+ },
95
+ {
96
+ "epoch": 0.13,
97
+ "grad_norm": 0.5055803656578064,
98
+ "learning_rate": 1.0072314049586778e-05,
99
+ "loss": 0.4264,
100
+ "step": 325
101
+ },
102
+ {
103
+ "epoch": 0.14,
104
+ "grad_norm": 0.5727549195289612,
105
+ "learning_rate": 1.084710743801653e-05,
106
+ "loss": 0.4373,
107
+ "step": 350
108
+ },
109
+ {
110
+ "epoch": 0.16,
111
+ "grad_norm": 0.6658156514167786,
112
+ "learning_rate": 1.1621900826446282e-05,
113
+ "loss": 0.3876,
114
+ "step": 375
115
+ },
116
+ {
117
+ "epoch": 0.17,
118
+ "grad_norm": 0.6771320700645447,
119
+ "learning_rate": 1.2396694214876034e-05,
120
+ "loss": 0.3518,
121
+ "step": 400
122
+ },
123
+ {
124
+ "epoch": 0.18,
125
+ "grad_norm": 0.6591110229492188,
126
+ "learning_rate": 1.3171487603305787e-05,
127
+ "loss": 0.3193,
128
+ "step": 425
129
+ },
130
+ {
131
+ "epoch": 0.19,
132
+ "grad_norm": 0.8635509610176086,
133
+ "learning_rate": 1.3946280991735537e-05,
134
+ "loss": 0.3055,
135
+ "step": 450
136
+ },
137
+ {
138
+ "epoch": 0.2,
139
+ "grad_norm": 1.0686248540878296,
140
+ "learning_rate": 1.472107438016529e-05,
141
+ "loss": 0.2824,
142
+ "step": 475
143
+ },
144
+ {
145
+ "epoch": 0.21,
146
+ "grad_norm": 0.88297039270401,
147
+ "learning_rate": 1.549586776859504e-05,
148
+ "loss": 0.2704,
149
+ "step": 500
150
+ },
151
+ {
152
+ "epoch": 0.22,
153
+ "grad_norm": 1.0282652378082275,
154
+ "learning_rate": 1.6270661157024794e-05,
155
+ "loss": 0.2384,
156
+ "step": 525
157
+ },
158
+ {
159
+ "epoch": 0.23,
160
+ "grad_norm": 1.009024977684021,
161
+ "learning_rate": 1.7045454545454546e-05,
162
+ "loss": 0.2241,
163
+ "step": 550
164
+ },
165
+ {
166
+ "epoch": 0.24,
167
+ "grad_norm": 1.1481122970581055,
168
+ "learning_rate": 1.78202479338843e-05,
169
+ "loss": 0.1914,
170
+ "step": 575
171
+ },
172
+ {
173
+ "epoch": 0.25,
174
+ "grad_norm": 1.3241642713546753,
175
+ "learning_rate": 1.859504132231405e-05,
176
+ "loss": 0.1672,
177
+ "step": 600
178
+ },
179
+ {
180
+ "epoch": 0.26,
181
+ "grad_norm": 1.0569857358932495,
182
+ "learning_rate": 1.9369834710743803e-05,
183
+ "loss": 0.1936,
184
+ "step": 625
185
+ },
186
+ {
187
+ "epoch": 0.27,
188
+ "grad_norm": 0.8776970505714417,
189
+ "learning_rate": 2.0144628099173555e-05,
190
+ "loss": 0.1528,
191
+ "step": 650
192
+ },
193
+ {
194
+ "epoch": 0.28,
195
+ "grad_norm": 0.7247719168663025,
196
+ "learning_rate": 2.0919421487603307e-05,
197
+ "loss": 0.1635,
198
+ "step": 675
199
+ },
200
+ {
201
+ "epoch": 0.29,
202
+ "grad_norm": 1.1265312433242798,
203
+ "learning_rate": 2.169421487603306e-05,
204
+ "loss": 0.1432,
205
+ "step": 700
206
+ },
207
+ {
208
+ "epoch": 0.3,
209
+ "grad_norm": 0.9996057748794556,
210
+ "learning_rate": 2.2469008264462812e-05,
211
+ "loss": 0.1448,
212
+ "step": 725
213
+ },
214
+ {
215
+ "epoch": 0.31,
216
+ "grad_norm": 0.7036225199699402,
217
+ "learning_rate": 2.3243801652892564e-05,
218
+ "loss": 0.1505,
219
+ "step": 750
220
+ },
221
+ {
222
+ "epoch": 0.32,
223
+ "grad_norm": 0.8441881537437439,
224
+ "learning_rate": 2.4018595041322316e-05,
225
+ "loss": 0.1446,
226
+ "step": 775
227
+ },
228
+ {
229
+ "epoch": 0.33,
230
+ "grad_norm": 1.1108651161193848,
231
+ "learning_rate": 2.479338842975207e-05,
232
+ "loss": 0.1291,
233
+ "step": 800
234
+ },
235
+ {
236
+ "epoch": 0.34,
237
+ "grad_norm": 1.3914504051208496,
238
+ "learning_rate": 2.556818181818182e-05,
239
+ "loss": 0.1188,
240
+ "step": 825
241
+ },
242
+ {
243
+ "epoch": 0.35,
244
+ "grad_norm": 1.5808395147323608,
245
+ "learning_rate": 2.6342975206611573e-05,
246
+ "loss": 0.1294,
247
+ "step": 850
248
+ },
249
+ {
250
+ "epoch": 0.36,
251
+ "grad_norm": 0.8349003195762634,
252
+ "learning_rate": 2.7117768595041322e-05,
253
+ "loss": 0.1396,
254
+ "step": 875
255
+ },
256
+ {
257
+ "epoch": 0.37,
258
+ "grad_norm": 0.9605033993721008,
259
+ "learning_rate": 2.7892561983471074e-05,
260
+ "loss": 0.1229,
261
+ "step": 900
262
+ },
263
+ {
264
+ "epoch": 0.38,
265
+ "grad_norm": 0.8516783118247986,
266
+ "learning_rate": 2.8667355371900826e-05,
267
+ "loss": 0.116,
268
+ "step": 925
269
+ },
270
+ {
271
+ "epoch": 0.39,
272
+ "grad_norm": 1.3230878114700317,
273
+ "learning_rate": 2.944214876033058e-05,
274
+ "loss": 0.1224,
275
+ "step": 950
276
+ },
277
+ {
278
+ "epoch": 0.4,
279
+ "grad_norm": 0.7461815476417542,
280
+ "learning_rate": 2.9975884244372988e-05,
281
+ "loss": 0.119,
282
+ "step": 975
283
+ },
284
+ {
285
+ "epoch": 0.41,
286
+ "grad_norm": 0.6247888803482056,
287
+ "learning_rate": 2.98897565457051e-05,
288
+ "loss": 0.0974,
289
+ "step": 1000
290
+ },
291
+ {
292
+ "epoch": 0.42,
293
+ "grad_norm": 0.6625025868415833,
294
+ "learning_rate": 2.980362884703721e-05,
295
+ "loss": 0.111,
296
+ "step": 1025
297
+ },
298
+ {
299
+ "epoch": 0.43,
300
+ "grad_norm": 0.8011340498924255,
301
+ "learning_rate": 2.9717501148369314e-05,
302
+ "loss": 0.1013,
303
+ "step": 1050
304
+ },
305
+ {
306
+ "epoch": 0.44,
307
+ "grad_norm": 0.6088280081748962,
308
+ "learning_rate": 2.9631373449701425e-05,
309
+ "loss": 0.1135,
310
+ "step": 1075
311
+ },
312
+ {
313
+ "epoch": 0.45,
314
+ "grad_norm": 0.32283875346183777,
315
+ "learning_rate": 2.9545245751033532e-05,
316
+ "loss": 0.1039,
317
+ "step": 1100
318
+ },
319
+ {
320
+ "epoch": 0.47,
321
+ "grad_norm": 0.8527675271034241,
322
+ "learning_rate": 2.945911805236564e-05,
323
+ "loss": 0.1119,
324
+ "step": 1125
325
+ },
326
+ {
327
+ "epoch": 0.48,
328
+ "grad_norm": 0.6812451481819153,
329
+ "learning_rate": 2.937299035369775e-05,
330
+ "loss": 0.0999,
331
+ "step": 1150
332
+ },
333
+ {
334
+ "epoch": 0.49,
335
+ "grad_norm": 0.3629147708415985,
336
+ "learning_rate": 2.928686265502986e-05,
337
+ "loss": 0.0947,
338
+ "step": 1175
339
+ },
340
+ {
341
+ "epoch": 0.5,
342
+ "grad_norm": 0.9273631572723389,
343
+ "learning_rate": 2.9200734956361966e-05,
344
+ "loss": 0.1101,
345
+ "step": 1200
346
+ },
347
+ {
348
+ "epoch": 0.51,
349
+ "grad_norm": 0.4564916491508484,
350
+ "learning_rate": 2.9114607257694073e-05,
351
+ "loss": 0.1019,
352
+ "step": 1225
353
+ },
354
+ {
355
+ "epoch": 0.52,
356
+ "grad_norm": 1.147258996963501,
357
+ "learning_rate": 2.9028479559026184e-05,
358
+ "loss": 0.0865,
359
+ "step": 1250
360
+ },
361
+ {
362
+ "epoch": 0.53,
363
+ "grad_norm": 0.4461434781551361,
364
+ "learning_rate": 2.8942351860358292e-05,
365
+ "loss": 0.0925,
366
+ "step": 1275
367
+ },
368
+ {
369
+ "epoch": 0.54,
370
+ "grad_norm": 0.5906102657318115,
371
+ "learning_rate": 2.88562241616904e-05,
372
+ "loss": 0.1044,
373
+ "step": 1300
374
+ },
375
+ {
376
+ "epoch": 0.55,
377
+ "grad_norm": 0.44959595799446106,
378
+ "learning_rate": 2.877009646302251e-05,
379
+ "loss": 0.1008,
380
+ "step": 1325
381
+ },
382
+ {
383
+ "epoch": 0.56,
384
+ "grad_norm": 0.918072521686554,
385
+ "learning_rate": 2.8683968764354614e-05,
386
+ "loss": 0.0904,
387
+ "step": 1350
388
+ },
389
+ {
390
+ "epoch": 0.57,
391
+ "grad_norm": 0.47751984000205994,
392
+ "learning_rate": 2.8597841065686725e-05,
393
+ "loss": 0.0821,
394
+ "step": 1375
395
+ },
396
+ {
397
+ "epoch": 0.58,
398
+ "grad_norm": 0.37669825553894043,
399
+ "learning_rate": 2.8511713367018833e-05,
400
+ "loss": 0.0889,
401
+ "step": 1400
402
+ },
403
+ {
404
+ "epoch": 0.59,
405
+ "grad_norm": 0.5528315305709839,
406
+ "learning_rate": 2.842558566835094e-05,
407
+ "loss": 0.0884,
408
+ "step": 1425
409
+ },
410
+ {
411
+ "epoch": 0.6,
412
+ "grad_norm": 0.7219104170799255,
413
+ "learning_rate": 2.833945796968305e-05,
414
+ "loss": 0.0909,
415
+ "step": 1450
416
+ },
417
+ {
418
+ "epoch": 0.61,
419
+ "grad_norm": 0.5634051561355591,
420
+ "learning_rate": 2.825333027101516e-05,
421
+ "loss": 0.0868,
422
+ "step": 1475
423
+ },
424
+ {
425
+ "epoch": 0.62,
426
+ "grad_norm": 0.33194178342819214,
427
+ "learning_rate": 2.816720257234727e-05,
428
+ "loss": 0.0901,
429
+ "step": 1500
430
+ },
431
+ {
432
+ "epoch": 0.63,
433
+ "grad_norm": 0.6954057812690735,
434
+ "learning_rate": 2.8081074873679374e-05,
435
+ "loss": 0.0834,
436
+ "step": 1525
437
+ },
438
+ {
439
+ "epoch": 0.64,
440
+ "grad_norm": 0.5542522072792053,
441
+ "learning_rate": 2.7994947175011485e-05,
442
+ "loss": 0.0923,
443
+ "step": 1550
444
+ },
445
+ {
446
+ "epoch": 0.65,
447
+ "grad_norm": 0.5636582970619202,
448
+ "learning_rate": 2.7908819476343596e-05,
449
+ "loss": 0.0887,
450
+ "step": 1575
451
+ },
452
+ {
453
+ "epoch": 0.66,
454
+ "grad_norm": 0.6528864502906799,
455
+ "learning_rate": 2.78226917776757e-05,
456
+ "loss": 0.0776,
457
+ "step": 1600
458
+ },
459
+ {
460
+ "epoch": 0.67,
461
+ "grad_norm": 0.3220342695713043,
462
+ "learning_rate": 2.773656407900781e-05,
463
+ "loss": 0.0813,
464
+ "step": 1625
465
+ },
466
+ {
467
+ "epoch": 0.68,
468
+ "grad_norm": 0.4045586884021759,
469
+ "learning_rate": 2.765043638033992e-05,
470
+ "loss": 0.0888,
471
+ "step": 1650
472
+ },
473
+ {
474
+ "epoch": 0.69,
475
+ "grad_norm": 0.7669140100479126,
476
+ "learning_rate": 2.7564308681672026e-05,
477
+ "loss": 0.0949,
478
+ "step": 1675
479
+ },
480
+ {
481
+ "epoch": 0.7,
482
+ "grad_norm": 0.3345740735530853,
483
+ "learning_rate": 2.7478180983004137e-05,
484
+ "loss": 0.0876,
485
+ "step": 1700
486
+ },
487
+ {
488
+ "epoch": 0.71,
489
+ "grad_norm": 0.7757517695426941,
490
+ "learning_rate": 2.7392053284336244e-05,
491
+ "loss": 0.0799,
492
+ "step": 1725
493
+ },
494
+ {
495
+ "epoch": 0.72,
496
+ "grad_norm": 0.5490168929100037,
497
+ "learning_rate": 2.7305925585668352e-05,
498
+ "loss": 0.0814,
499
+ "step": 1750
500
+ },
501
+ {
502
+ "epoch": 0.73,
503
+ "grad_norm": 0.4781877398490906,
504
+ "learning_rate": 2.721979788700046e-05,
505
+ "loss": 0.0785,
506
+ "step": 1775
507
+ },
508
+ {
509
+ "epoch": 0.74,
510
+ "grad_norm": 0.5043837428092957,
511
+ "learning_rate": 2.713367018833257e-05,
512
+ "loss": 0.0834,
513
+ "step": 1800
514
+ },
515
+ {
516
+ "epoch": 0.75,
517
+ "grad_norm": 0.4992827773094177,
518
+ "learning_rate": 2.7047542489664674e-05,
519
+ "loss": 0.0921,
520
+ "step": 1825
521
+ },
522
+ {
523
+ "epoch": 0.76,
524
+ "grad_norm": 0.6074797511100769,
525
+ "learning_rate": 2.6961414790996785e-05,
526
+ "loss": 0.0909,
527
+ "step": 1850
528
+ },
529
+ {
530
+ "epoch": 0.78,
531
+ "grad_norm": 0.6552305817604065,
532
+ "learning_rate": 2.6875287092328896e-05,
533
+ "loss": 0.0784,
534
+ "step": 1875
535
+ },
536
+ {
537
+ "epoch": 0.79,
538
+ "grad_norm": 0.5910527110099792,
539
+ "learning_rate": 2.6789159393661e-05,
540
+ "loss": 0.0848,
541
+ "step": 1900
542
+ },
543
+ {
544
+ "epoch": 0.8,
545
+ "grad_norm": 0.7094011306762695,
546
+ "learning_rate": 2.670303169499311e-05,
547
+ "loss": 0.0857,
548
+ "step": 1925
549
+ },
550
+ {
551
+ "epoch": 0.81,
552
+ "grad_norm": 0.6078013777732849,
553
+ "learning_rate": 2.661690399632522e-05,
554
+ "loss": 0.0869,
555
+ "step": 1950
556
+ },
557
+ {
558
+ "epoch": 0.82,
559
+ "grad_norm": 0.43232443928718567,
560
+ "learning_rate": 2.6530776297657326e-05,
561
+ "loss": 0.0834,
562
+ "step": 1975
563
+ },
564
+ {
565
+ "epoch": 0.83,
566
+ "grad_norm": 0.4060278534889221,
567
+ "learning_rate": 2.6444648598989437e-05,
568
+ "loss": 0.0813,
569
+ "step": 2000
570
+ },
571
+ {
572
+ "epoch": 0.84,
573
+ "grad_norm": 0.6071100234985352,
574
+ "learning_rate": 2.6358520900321545e-05,
575
+ "loss": 0.0795,
576
+ "step": 2025
577
+ },
578
+ {
579
+ "epoch": 0.85,
580
+ "grad_norm": 0.41472747921943665,
581
+ "learning_rate": 2.6272393201653652e-05,
582
+ "loss": 0.0828,
583
+ "step": 2050
584
+ },
585
+ {
586
+ "epoch": 0.86,
587
+ "grad_norm": 0.586531400680542,
588
+ "learning_rate": 2.618626550298576e-05,
589
+ "loss": 0.0736,
590
+ "step": 2075
591
+ },
592
+ {
593
+ "epoch": 0.87,
594
+ "grad_norm": 0.42143017053604126,
595
+ "learning_rate": 2.610013780431787e-05,
596
+ "loss": 0.078,
597
+ "step": 2100
598
+ },
599
+ {
600
+ "epoch": 0.88,
601
+ "grad_norm": 0.35077112913131714,
602
+ "learning_rate": 2.601401010564998e-05,
603
+ "loss": 0.0844,
604
+ "step": 2125
605
+ },
606
+ {
607
+ "epoch": 0.89,
608
+ "grad_norm": 0.5425326824188232,
609
+ "learning_rate": 2.5927882406982086e-05,
610
+ "loss": 0.0772,
611
+ "step": 2150
612
+ },
613
+ {
614
+ "epoch": 0.9,
615
+ "grad_norm": 0.25989875197410583,
616
+ "learning_rate": 2.5841754708314197e-05,
617
+ "loss": 0.0749,
618
+ "step": 2175
619
+ },
620
+ {
621
+ "epoch": 0.91,
622
+ "grad_norm": 0.3507814407348633,
623
+ "learning_rate": 2.57556270096463e-05,
624
+ "loss": 0.0859,
625
+ "step": 2200
626
+ },
627
+ {
628
+ "epoch": 0.92,
629
+ "grad_norm": 0.39133837819099426,
630
+ "learning_rate": 2.5669499310978412e-05,
631
+ "loss": 0.0765,
632
+ "step": 2225
633
+ },
634
+ {
635
+ "epoch": 0.93,
636
+ "grad_norm": 0.3232302963733673,
637
+ "learning_rate": 2.558337161231052e-05,
638
+ "loss": 0.0737,
639
+ "step": 2250
640
+ },
641
+ {
642
+ "epoch": 0.94,
643
+ "grad_norm": 1.1417970657348633,
644
+ "learning_rate": 2.5497243913642627e-05,
645
+ "loss": 0.0796,
646
+ "step": 2275
647
+ },
648
+ {
649
+ "epoch": 0.95,
650
+ "grad_norm": 0.6268133521080017,
651
+ "learning_rate": 2.5411116214974738e-05,
652
+ "loss": 0.0723,
653
+ "step": 2300
654
+ },
655
+ {
656
+ "epoch": 0.96,
657
+ "grad_norm": 0.4392102360725403,
658
+ "learning_rate": 2.5324988516306845e-05,
659
+ "loss": 0.0742,
660
+ "step": 2325
661
+ },
662
+ {
663
+ "epoch": 0.97,
664
+ "grad_norm": 0.3616221249103546,
665
+ "learning_rate": 2.5238860817638953e-05,
666
+ "loss": 0.0811,
667
+ "step": 2350
668
+ },
669
+ {
670
+ "epoch": 0.98,
671
+ "grad_norm": 0.5831270813941956,
672
+ "learning_rate": 2.515273311897106e-05,
673
+ "loss": 0.0741,
674
+ "step": 2375
675
+ },
676
+ {
677
+ "epoch": 0.99,
678
+ "grad_norm": 0.4768739640712738,
679
+ "learning_rate": 2.506660542030317e-05,
680
+ "loss": 0.0813,
681
+ "step": 2400
682
+ },
683
+ {
684
+ "epoch": 1.0,
685
+ "grad_norm": 0.3268464207649231,
686
+ "learning_rate": 2.498047772163528e-05,
687
+ "loss": 0.061,
688
+ "step": 2425
689
+ },
690
+ {
691
+ "epoch": 1.01,
692
+ "grad_norm": 0.5923863649368286,
693
+ "learning_rate": 2.4894350022967386e-05,
694
+ "loss": 0.0799,
695
+ "step": 2450
696
+ },
697
+ {
698
+ "epoch": 1.02,
699
+ "grad_norm": 0.3295147120952606,
700
+ "learning_rate": 2.4808222324299497e-05,
701
+ "loss": 0.0742,
702
+ "step": 2475
703
+ },
704
+ {
705
+ "epoch": 1.03,
706
+ "grad_norm": 0.6255597472190857,
707
+ "learning_rate": 2.47220946256316e-05,
708
+ "loss": 0.0685,
709
+ "step": 2500
710
+ },
711
+ {
712
+ "epoch": 1.04,
713
+ "grad_norm": 0.28953835368156433,
714
+ "learning_rate": 2.4635966926963712e-05,
715
+ "loss": 0.0645,
716
+ "step": 2525
717
+ },
718
+ {
719
+ "epoch": 1.05,
720
+ "grad_norm": 0.28053247928619385,
721
+ "learning_rate": 2.4549839228295823e-05,
722
+ "loss": 0.0756,
723
+ "step": 2550
724
+ },
725
+ {
726
+ "epoch": 1.06,
727
+ "grad_norm": 0.48497211933135986,
728
+ "learning_rate": 2.4463711529627927e-05,
729
+ "loss": 0.0659,
730
+ "step": 2575
731
+ },
732
+ {
733
+ "epoch": 1.07,
734
+ "grad_norm": 0.4199415445327759,
735
+ "learning_rate": 2.437758383096004e-05,
736
+ "loss": 0.0783,
737
+ "step": 2600
738
+ },
739
+ {
740
+ "epoch": 1.09,
741
+ "grad_norm": 0.5894821882247925,
742
+ "learning_rate": 2.4291456132292146e-05,
743
+ "loss": 0.0655,
744
+ "step": 2625
745
+ },
746
+ {
747
+ "epoch": 1.1,
748
+ "grad_norm": 0.31835877895355225,
749
+ "learning_rate": 2.4205328433624253e-05,
750
+ "loss": 0.0595,
751
+ "step": 2650
752
+ },
753
+ {
754
+ "epoch": 1.11,
755
+ "grad_norm": 0.3910796642303467,
756
+ "learning_rate": 2.411920073495636e-05,
757
+ "loss": 0.0703,
758
+ "step": 2675
759
+ },
760
+ {
761
+ "epoch": 1.12,
762
+ "grad_norm": 0.4052046239376068,
763
+ "learning_rate": 2.4033073036288472e-05,
764
+ "loss": 0.0751,
765
+ "step": 2700
766
+ },
767
+ {
768
+ "epoch": 1.13,
769
+ "grad_norm": 0.5519680380821228,
770
+ "learning_rate": 2.394694533762058e-05,
771
+ "loss": 0.0643,
772
+ "step": 2725
773
+ },
774
+ {
775
+ "epoch": 1.14,
776
+ "grad_norm": 0.2853243052959442,
777
+ "learning_rate": 2.3860817638952687e-05,
778
+ "loss": 0.0582,
779
+ "step": 2750
780
+ },
781
+ {
782
+ "epoch": 1.15,
783
+ "grad_norm": 0.3067447245121002,
784
+ "learning_rate": 2.3774689940284798e-05,
785
+ "loss": 0.0676,
786
+ "step": 2775
787
+ },
788
+ {
789
+ "epoch": 1.16,
790
+ "grad_norm": 0.3290148079395294,
791
+ "learning_rate": 2.3688562241616902e-05,
792
+ "loss": 0.0679,
793
+ "step": 2800
794
+ },
795
+ {
796
+ "epoch": 1.17,
797
+ "grad_norm": 0.3863273561000824,
798
+ "learning_rate": 2.3602434542949013e-05,
799
+ "loss": 0.0606,
800
+ "step": 2825
801
+ },
802
+ {
803
+ "epoch": 1.18,
804
+ "grad_norm": 0.4047059416770935,
805
+ "learning_rate": 2.3516306844281124e-05,
806
+ "loss": 0.0668,
807
+ "step": 2850
808
+ },
809
+ {
810
+ "epoch": 1.19,
811
+ "grad_norm": 0.42530855536460876,
812
+ "learning_rate": 2.3430179145613228e-05,
813
+ "loss": 0.0606,
814
+ "step": 2875
815
+ },
816
+ {
817
+ "epoch": 1.2,
818
+ "grad_norm": 0.3491179943084717,
819
+ "learning_rate": 2.334405144694534e-05,
820
+ "loss": 0.0762,
821
+ "step": 2900
822
+ },
823
+ {
824
+ "epoch": 1.21,
825
+ "grad_norm": 0.15498507022857666,
826
+ "learning_rate": 2.3257923748277446e-05,
827
+ "loss": 0.0719,
828
+ "step": 2925
829
+ },
830
+ {
831
+ "epoch": 1.22,
832
+ "grad_norm": 0.4127048850059509,
833
+ "learning_rate": 2.3171796049609554e-05,
834
+ "loss": 0.0614,
835
+ "step": 2950
836
+ },
837
+ {
838
+ "epoch": 1.23,
839
+ "grad_norm": 0.3426351249217987,
840
+ "learning_rate": 2.3085668350941665e-05,
841
+ "loss": 0.0624,
842
+ "step": 2975
843
+ },
844
+ {
845
+ "epoch": 1.24,
846
+ "grad_norm": 0.28411659598350525,
847
+ "learning_rate": 2.2999540652273772e-05,
848
+ "loss": 0.0662,
849
+ "step": 3000
850
+ },
851
+ {
852
+ "epoch": 1.25,
853
+ "grad_norm": 0.16444465517997742,
854
+ "learning_rate": 2.291341295360588e-05,
855
+ "loss": 0.0615,
856
+ "step": 3025
857
+ },
858
+ {
859
+ "epoch": 1.26,
860
+ "grad_norm": 0.3092861771583557,
861
+ "learning_rate": 2.2827285254937987e-05,
862
+ "loss": 0.0582,
863
+ "step": 3050
864
+ },
865
+ {
866
+ "epoch": 1.27,
867
+ "grad_norm": 0.483084499835968,
868
+ "learning_rate": 2.2741157556270098e-05,
869
+ "loss": 0.0595,
870
+ "step": 3075
871
+ },
872
+ {
873
+ "epoch": 1.28,
874
+ "grad_norm": 0.2572250962257385,
875
+ "learning_rate": 2.2655029857602202e-05,
876
+ "loss": 0.0699,
877
+ "step": 3100
878
+ },
879
+ {
880
+ "epoch": 1.29,
881
+ "grad_norm": 0.21841402351856232,
882
+ "learning_rate": 2.2568902158934313e-05,
883
+ "loss": 0.0606,
884
+ "step": 3125
885
+ },
886
+ {
887
+ "epoch": 1.3,
888
+ "grad_norm": 0.33092397451400757,
889
+ "learning_rate": 2.2482774460266424e-05,
890
+ "loss": 0.0649,
891
+ "step": 3150
892
+ },
893
+ {
894
+ "epoch": 1.31,
895
+ "grad_norm": 0.35950547456741333,
896
+ "learning_rate": 2.239664676159853e-05,
897
+ "loss": 0.0644,
898
+ "step": 3175
899
+ },
900
+ {
901
+ "epoch": 1.32,
902
+ "grad_norm": 0.33506807684898376,
903
+ "learning_rate": 2.231051906293064e-05,
904
+ "loss": 0.0662,
905
+ "step": 3200
906
+ },
907
+ {
908
+ "epoch": 1.33,
909
+ "grad_norm": 0.5829468369483948,
910
+ "learning_rate": 2.2224391364262747e-05,
911
+ "loss": 0.0643,
912
+ "step": 3225
913
+ },
914
+ {
915
+ "epoch": 1.34,
916
+ "grad_norm": 0.3790678381919861,
917
+ "learning_rate": 2.2138263665594854e-05,
918
+ "loss": 0.0729,
919
+ "step": 3250
920
+ },
921
+ {
922
+ "epoch": 1.35,
923
+ "grad_norm": 0.34478724002838135,
924
+ "learning_rate": 2.2052135966926965e-05,
925
+ "loss": 0.0645,
926
+ "step": 3275
927
+ },
928
+ {
929
+ "epoch": 1.36,
930
+ "grad_norm": 0.3655211925506592,
931
+ "learning_rate": 2.1966008268259073e-05,
932
+ "loss": 0.0756,
933
+ "step": 3300
934
+ },
935
+ {
936
+ "epoch": 1.37,
937
+ "grad_norm": 0.3092879354953766,
938
+ "learning_rate": 2.187988056959118e-05,
939
+ "loss": 0.0677,
940
+ "step": 3325
941
+ },
942
+ {
943
+ "epoch": 1.38,
944
+ "grad_norm": 0.3613986372947693,
945
+ "learning_rate": 2.1793752870923288e-05,
946
+ "loss": 0.0627,
947
+ "step": 3350
948
+ },
949
+ {
950
+ "epoch": 1.4,
951
+ "grad_norm": 0.6292563080787659,
952
+ "learning_rate": 2.17076251722554e-05,
953
+ "loss": 0.0698,
954
+ "step": 3375
955
+ },
956
+ {
957
+ "epoch": 1.41,
958
+ "grad_norm": 0.24357610940933228,
959
+ "learning_rate": 2.1621497473587506e-05,
960
+ "loss": 0.0701,
961
+ "step": 3400
962
+ },
963
+ {
964
+ "epoch": 1.42,
965
+ "grad_norm": 0.32474613189697266,
966
+ "learning_rate": 2.1535369774919614e-05,
967
+ "loss": 0.0662,
968
+ "step": 3425
969
+ },
970
+ {
971
+ "epoch": 1.43,
972
+ "grad_norm": 0.30164840817451477,
973
+ "learning_rate": 2.1449242076251725e-05,
974
+ "loss": 0.0613,
975
+ "step": 3450
976
+ },
977
+ {
978
+ "epoch": 1.44,
979
+ "grad_norm": 0.25660645961761475,
980
+ "learning_rate": 2.136311437758383e-05,
981
+ "loss": 0.0622,
982
+ "step": 3475
983
+ },
984
+ {
985
+ "epoch": 1.45,
986
+ "grad_norm": 0.4381033480167389,
987
+ "learning_rate": 2.127698667891594e-05,
988
+ "loss": 0.0676,
989
+ "step": 3500
990
+ },
991
+ {
992
+ "epoch": 1.46,
993
+ "grad_norm": 0.36808228492736816,
994
+ "learning_rate": 2.1190858980248047e-05,
995
+ "loss": 0.0665,
996
+ "step": 3525
997
+ },
998
+ {
999
+ "epoch": 1.47,
1000
+ "grad_norm": 0.2712726294994354,
1001
+ "learning_rate": 2.1104731281580155e-05,
1002
+ "loss": 0.0728,
1003
+ "step": 3550
1004
+ },
1005
+ {
1006
+ "epoch": 1.48,
1007
+ "grad_norm": 0.5337083339691162,
1008
+ "learning_rate": 2.1018603582912266e-05,
1009
+ "loss": 0.0666,
1010
+ "step": 3575
1011
+ },
1012
+ {
1013
+ "epoch": 1.49,
1014
+ "grad_norm": 0.33135494589805603,
1015
+ "learning_rate": 2.0932475884244373e-05,
1016
+ "loss": 0.0721,
1017
+ "step": 3600
1018
+ },
1019
+ {
1020
+ "epoch": 1.5,
1021
+ "grad_norm": 0.3100278675556183,
1022
+ "learning_rate": 2.084634818557648e-05,
1023
+ "loss": 0.066,
1024
+ "step": 3625
1025
+ },
1026
+ {
1027
+ "epoch": 1.51,
1028
+ "grad_norm": 0.4423840045928955,
1029
+ "learning_rate": 2.076022048690859e-05,
1030
+ "loss": 0.0666,
1031
+ "step": 3650
1032
+ },
1033
+ {
1034
+ "epoch": 1.52,
1035
+ "grad_norm": 0.3579668700695038,
1036
+ "learning_rate": 2.06740927882407e-05,
1037
+ "loss": 0.0635,
1038
+ "step": 3675
1039
+ },
1040
+ {
1041
+ "epoch": 1.53,
1042
+ "grad_norm": 0.4105582535266876,
1043
+ "learning_rate": 2.0587965089572807e-05,
1044
+ "loss": 0.0683,
1045
+ "step": 3700
1046
+ },
1047
+ {
1048
+ "epoch": 1.54,
1049
+ "grad_norm": 0.5205901861190796,
1050
+ "learning_rate": 2.0501837390904914e-05,
1051
+ "loss": 0.0725,
1052
+ "step": 3725
1053
+ },
1054
+ {
1055
+ "epoch": 1.55,
1056
+ "grad_norm": 0.508314311504364,
1057
+ "learning_rate": 2.0415709692237025e-05,
1058
+ "loss": 0.0751,
1059
+ "step": 3750
1060
+ },
1061
+ {
1062
+ "epoch": 1.56,
1063
+ "grad_norm": 0.36034587025642395,
1064
+ "learning_rate": 2.032958199356913e-05,
1065
+ "loss": 0.0685,
1066
+ "step": 3775
1067
+ },
1068
+ {
1069
+ "epoch": 1.57,
1070
+ "grad_norm": 0.2791132926940918,
1071
+ "learning_rate": 2.024345429490124e-05,
1072
+ "loss": 0.0642,
1073
+ "step": 3800
1074
+ },
1075
+ {
1076
+ "epoch": 1.58,
1077
+ "grad_norm": 0.41801777482032776,
1078
+ "learning_rate": 2.015732659623335e-05,
1079
+ "loss": 0.0552,
1080
+ "step": 3825
1081
+ },
1082
+ {
1083
+ "epoch": 1.59,
1084
+ "grad_norm": 0.2085207998752594,
1085
+ "learning_rate": 2.0071198897565455e-05,
1086
+ "loss": 0.057,
1087
+ "step": 3850
1088
+ },
1089
+ {
1090
+ "epoch": 1.6,
1091
+ "grad_norm": 0.3542513847351074,
1092
+ "learning_rate": 1.9985071198897566e-05,
1093
+ "loss": 0.0635,
1094
+ "step": 3875
1095
+ },
1096
+ {
1097
+ "epoch": 1.61,
1098
+ "grad_norm": 0.21708066761493683,
1099
+ "learning_rate": 1.9898943500229674e-05,
1100
+ "loss": 0.0609,
1101
+ "step": 3900
1102
+ },
1103
+ {
1104
+ "epoch": 1.62,
1105
+ "grad_norm": 0.39065513014793396,
1106
+ "learning_rate": 1.9812815801561785e-05,
1107
+ "loss": 0.0638,
1108
+ "step": 3925
1109
+ },
1110
+ {
1111
+ "epoch": 1.63,
1112
+ "grad_norm": 0.29231196641921997,
1113
+ "learning_rate": 1.972668810289389e-05,
1114
+ "loss": 0.0665,
1115
+ "step": 3950
1116
+ },
1117
+ {
1118
+ "epoch": 1.64,
1119
+ "grad_norm": 0.24261677265167236,
1120
+ "learning_rate": 1.9640560404226e-05,
1121
+ "loss": 0.0621,
1122
+ "step": 3975
1123
+ },
1124
+ {
1125
+ "epoch": 1.65,
1126
+ "grad_norm": 0.3188820779323578,
1127
+ "learning_rate": 1.955443270555811e-05,
1128
+ "loss": 0.0579,
1129
+ "step": 4000
1130
+ },
1131
+ {
1132
+ "epoch": 1.66,
1133
+ "grad_norm": 0.44512245059013367,
1134
+ "learning_rate": 1.9468305006890215e-05,
1135
+ "loss": 0.0638,
1136
+ "step": 4025
1137
+ },
1138
+ {
1139
+ "epoch": 1.67,
1140
+ "grad_norm": 0.3457041084766388,
1141
+ "learning_rate": 1.9382177308222326e-05,
1142
+ "loss": 0.062,
1143
+ "step": 4050
1144
+ },
1145
+ {
1146
+ "epoch": 1.68,
1147
+ "grad_norm": 0.20665928721427917,
1148
+ "learning_rate": 1.9296049609554433e-05,
1149
+ "loss": 0.0614,
1150
+ "step": 4075
1151
+ },
1152
+ {
1153
+ "epoch": 1.69,
1154
+ "grad_norm": 0.3206002116203308,
1155
+ "learning_rate": 1.920992191088654e-05,
1156
+ "loss": 0.0714,
1157
+ "step": 4100
1158
+ },
1159
+ {
1160
+ "epoch": 1.71,
1161
+ "grad_norm": 0.2654290497303009,
1162
+ "learning_rate": 1.9123794212218652e-05,
1163
+ "loss": 0.0603,
1164
+ "step": 4125
1165
+ },
1166
+ {
1167
+ "epoch": 1.72,
1168
+ "grad_norm": 0.39113229513168335,
1169
+ "learning_rate": 1.903766651355076e-05,
1170
+ "loss": 0.0635,
1171
+ "step": 4150
1172
+ },
1173
+ {
1174
+ "epoch": 1.73,
1175
+ "grad_norm": 0.34004124999046326,
1176
+ "learning_rate": 1.8951538814882867e-05,
1177
+ "loss": 0.0638,
1178
+ "step": 4175
1179
+ },
1180
+ {
1181
+ "epoch": 1.74,
1182
+ "grad_norm": 0.27496448159217834,
1183
+ "learning_rate": 1.8865411116214974e-05,
1184
+ "loss": 0.0622,
1185
+ "step": 4200
1186
+ },
1187
+ {
1188
+ "epoch": 1.75,
1189
+ "grad_norm": 0.3293686509132385,
1190
+ "learning_rate": 1.8779283417547085e-05,
1191
+ "loss": 0.0652,
1192
+ "step": 4225
1193
+ },
1194
+ {
1195
+ "epoch": 1.76,
1196
+ "grad_norm": 0.2733684480190277,
1197
+ "learning_rate": 1.8693155718879193e-05,
1198
+ "loss": 0.0616,
1199
+ "step": 4250
1200
+ },
1201
+ {
1202
+ "epoch": 1.77,
1203
+ "grad_norm": 0.21654804050922394,
1204
+ "learning_rate": 1.86070280202113e-05,
1205
+ "loss": 0.0618,
1206
+ "step": 4275
1207
+ },
1208
+ {
1209
+ "epoch": 1.78,
1210
+ "grad_norm": 0.24511463940143585,
1211
+ "learning_rate": 1.852090032154341e-05,
1212
+ "loss": 0.0579,
1213
+ "step": 4300
1214
+ },
1215
+ {
1216
+ "epoch": 1.79,
1217
+ "grad_norm": 0.363006591796875,
1218
+ "learning_rate": 1.8434772622875515e-05,
1219
+ "loss": 0.0673,
1220
+ "step": 4325
1221
+ },
1222
+ {
1223
+ "epoch": 1.8,
1224
+ "grad_norm": 0.2865668535232544,
1225
+ "learning_rate": 1.8348644924207626e-05,
1226
+ "loss": 0.0633,
1227
+ "step": 4350
1228
+ },
1229
+ {
1230
+ "epoch": 1.81,
1231
+ "grad_norm": 0.20682819187641144,
1232
+ "learning_rate": 1.8262517225539734e-05,
1233
+ "loss": 0.0519,
1234
+ "step": 4375
1235
+ },
1236
+ {
1237
+ "epoch": 1.82,
1238
+ "grad_norm": 0.38699063658714294,
1239
+ "learning_rate": 1.817638952687184e-05,
1240
+ "loss": 0.0604,
1241
+ "step": 4400
1242
+ },
1243
+ {
1244
+ "epoch": 1.83,
1245
+ "grad_norm": 0.35452330112457275,
1246
+ "learning_rate": 1.8090261828203952e-05,
1247
+ "loss": 0.0568,
1248
+ "step": 4425
1249
+ },
1250
+ {
1251
+ "epoch": 1.84,
1252
+ "grad_norm": 0.2268667072057724,
1253
+ "learning_rate": 1.800413412953606e-05,
1254
+ "loss": 0.0653,
1255
+ "step": 4450
1256
+ },
1257
+ {
1258
+ "epoch": 1.85,
1259
+ "grad_norm": 0.30717945098876953,
1260
+ "learning_rate": 1.7918006430868167e-05,
1261
+ "loss": 0.0664,
1262
+ "step": 4475
1263
+ },
1264
+ {
1265
+ "epoch": 1.86,
1266
+ "grad_norm": 0.46489375829696655,
1267
+ "learning_rate": 1.7831878732200275e-05,
1268
+ "loss": 0.0562,
1269
+ "step": 4500
1270
+ },
1271
+ {
1272
+ "epoch": 1.87,
1273
+ "grad_norm": 0.43247151374816895,
1274
+ "learning_rate": 1.7745751033532386e-05,
1275
+ "loss": 0.0573,
1276
+ "step": 4525
1277
+ },
1278
+ {
1279
+ "epoch": 1.88,
1280
+ "grad_norm": 0.3387090563774109,
1281
+ "learning_rate": 1.7659623334864493e-05,
1282
+ "loss": 0.0519,
1283
+ "step": 4550
1284
+ },
1285
+ {
1286
+ "epoch": 1.89,
1287
+ "grad_norm": 0.16439248621463776,
1288
+ "learning_rate": 1.75734956361966e-05,
1289
+ "loss": 0.0663,
1290
+ "step": 4575
1291
+ },
1292
+ {
1293
+ "epoch": 1.9,
1294
+ "grad_norm": 0.8034338355064392,
1295
+ "learning_rate": 1.7487367937528712e-05,
1296
+ "loss": 0.0621,
1297
+ "step": 4600
1298
+ },
1299
+ {
1300
+ "epoch": 1.91,
1301
+ "grad_norm": 0.2516898512840271,
1302
+ "learning_rate": 1.7401240238860816e-05,
1303
+ "loss": 0.0612,
1304
+ "step": 4625
1305
+ },
1306
+ {
1307
+ "epoch": 1.92,
1308
+ "grad_norm": 0.22889916598796844,
1309
+ "learning_rate": 1.7315112540192927e-05,
1310
+ "loss": 0.0542,
1311
+ "step": 4650
1312
+ },
1313
+ {
1314
+ "epoch": 1.93,
1315
+ "grad_norm": 0.1372820883989334,
1316
+ "learning_rate": 1.7228984841525038e-05,
1317
+ "loss": 0.0591,
1318
+ "step": 4675
1319
+ },
1320
+ {
1321
+ "epoch": 1.94,
1322
+ "grad_norm": 0.3031134307384491,
1323
+ "learning_rate": 1.7142857142857142e-05,
1324
+ "loss": 0.0536,
1325
+ "step": 4700
1326
+ },
1327
+ {
1328
+ "epoch": 1.95,
1329
+ "grad_norm": 0.22743794322013855,
1330
+ "learning_rate": 1.7056729444189253e-05,
1331
+ "loss": 0.0579,
1332
+ "step": 4725
1333
+ },
1334
+ {
1335
+ "epoch": 1.96,
1336
+ "grad_norm": 0.419313907623291,
1337
+ "learning_rate": 1.697060174552136e-05,
1338
+ "loss": 0.0681,
1339
+ "step": 4750
1340
+ },
1341
+ {
1342
+ "epoch": 1.97,
1343
+ "grad_norm": 0.39816129207611084,
1344
+ "learning_rate": 1.6884474046853468e-05,
1345
+ "loss": 0.0629,
1346
+ "step": 4775
1347
+ },
1348
+ {
1349
+ "epoch": 1.98,
1350
+ "grad_norm": 0.3139801621437073,
1351
+ "learning_rate": 1.6798346348185575e-05,
1352
+ "loss": 0.0547,
1353
+ "step": 4800
1354
+ },
1355
+ {
1356
+ "epoch": 1.99,
1357
+ "grad_norm": 0.36901557445526123,
1358
+ "learning_rate": 1.6712218649517686e-05,
1359
+ "loss": 0.0631,
1360
+ "step": 4825
1361
+ },
1362
+ {
1363
+ "epoch": 2.0,
1364
+ "grad_norm": 0.3577054738998413,
1365
+ "learning_rate": 1.6626090950849794e-05,
1366
+ "loss": 0.0625,
1367
+ "step": 4850
1368
+ },
1369
+ {
1370
+ "epoch": 2.02,
1371
+ "grad_norm": 0.28602612018585205,
1372
+ "learning_rate": 1.65399632521819e-05,
1373
+ "loss": 0.0544,
1374
+ "step": 4875
1375
+ },
1376
+ {
1377
+ "epoch": 2.03,
1378
+ "grad_norm": 0.3862529397010803,
1379
+ "learning_rate": 1.6453835553514012e-05,
1380
+ "loss": 0.0544,
1381
+ "step": 4900
1382
+ },
1383
+ {
1384
+ "epoch": 2.04,
1385
+ "grad_norm": 0.2228946089744568,
1386
+ "learning_rate": 1.6367707854846116e-05,
1387
+ "loss": 0.0558,
1388
+ "step": 4925
1389
+ },
1390
+ {
1391
+ "epoch": 2.05,
1392
+ "grad_norm": 0.2535618841648102,
1393
+ "learning_rate": 1.6281580156178227e-05,
1394
+ "loss": 0.0521,
1395
+ "step": 4950
1396
+ },
1397
+ {
1398
+ "epoch": 2.06,
1399
+ "grad_norm": 0.22719667851924896,
1400
+ "learning_rate": 1.6195452457510338e-05,
1401
+ "loss": 0.056,
1402
+ "step": 4975
1403
+ },
1404
+ {
1405
+ "epoch": 2.07,
1406
+ "grad_norm": 0.3262714445590973,
1407
+ "learning_rate": 1.6109324758842442e-05,
1408
+ "loss": 0.0512,
1409
+ "step": 5000
1410
+ },
1411
+ {
1412
+ "epoch": 2.08,
1413
+ "grad_norm": 0.2454172968864441,
1414
+ "learning_rate": 1.6023197060174553e-05,
1415
+ "loss": 0.0499,
1416
+ "step": 5025
1417
+ },
1418
+ {
1419
+ "epoch": 2.09,
1420
+ "grad_norm": 0.6654460430145264,
1421
+ "learning_rate": 1.593706936150666e-05,
1422
+ "loss": 0.0516,
1423
+ "step": 5050
1424
+ },
1425
+ {
1426
+ "epoch": 2.1,
1427
+ "grad_norm": 0.37515684962272644,
1428
+ "learning_rate": 1.585094166283877e-05,
1429
+ "loss": 0.0487,
1430
+ "step": 5075
1431
+ },
1432
+ {
1433
+ "epoch": 2.11,
1434
+ "grad_norm": 0.3624023199081421,
1435
+ "learning_rate": 1.576481396417088e-05,
1436
+ "loss": 0.051,
1437
+ "step": 5100
1438
+ },
1439
+ {
1440
+ "epoch": 2.12,
1441
+ "grad_norm": 0.33498457074165344,
1442
+ "learning_rate": 1.5678686265502987e-05,
1443
+ "loss": 0.0489,
1444
+ "step": 5125
1445
+ },
1446
+ {
1447
+ "epoch": 2.13,
1448
+ "grad_norm": 0.4283454418182373,
1449
+ "learning_rate": 1.5592558566835094e-05,
1450
+ "loss": 0.0497,
1451
+ "step": 5150
1452
+ },
1453
+ {
1454
+ "epoch": 2.14,
1455
+ "grad_norm": 0.4568246006965637,
1456
+ "learning_rate": 1.5506430868167202e-05,
1457
+ "loss": 0.0552,
1458
+ "step": 5175
1459
+ },
1460
+ {
1461
+ "epoch": 2.15,
1462
+ "grad_norm": 0.3240622580051422,
1463
+ "learning_rate": 1.5420303169499313e-05,
1464
+ "loss": 0.0524,
1465
+ "step": 5200
1466
+ },
1467
+ {
1468
+ "epoch": 2.16,
1469
+ "grad_norm": 0.3433873951435089,
1470
+ "learning_rate": 1.533417547083142e-05,
1471
+ "loss": 0.0562,
1472
+ "step": 5225
1473
+ },
1474
+ {
1475
+ "epoch": 2.17,
1476
+ "grad_norm": 0.28902968764305115,
1477
+ "learning_rate": 1.5248047772163528e-05,
1478
+ "loss": 0.0564,
1479
+ "step": 5250
1480
+ },
1481
+ {
1482
+ "epoch": 2.18,
1483
+ "grad_norm": 0.5900773406028748,
1484
+ "learning_rate": 1.5161920073495637e-05,
1485
+ "loss": 0.0593,
1486
+ "step": 5275
1487
+ },
1488
+ {
1489
+ "epoch": 2.19,
1490
+ "grad_norm": 0.37150174379348755,
1491
+ "learning_rate": 1.5075792374827745e-05,
1492
+ "loss": 0.05,
1493
+ "step": 5300
1494
+ },
1495
+ {
1496
+ "epoch": 2.2,
1497
+ "grad_norm": 0.2919517159461975,
1498
+ "learning_rate": 1.4989664676159854e-05,
1499
+ "loss": 0.057,
1500
+ "step": 5325
1501
+ },
1502
+ {
1503
+ "epoch": 2.21,
1504
+ "grad_norm": 0.24746154248714447,
1505
+ "learning_rate": 1.4903536977491961e-05,
1506
+ "loss": 0.0549,
1507
+ "step": 5350
1508
+ },
1509
+ {
1510
+ "epoch": 2.22,
1511
+ "grad_norm": 0.4665452539920807,
1512
+ "learning_rate": 1.481740927882407e-05,
1513
+ "loss": 0.051,
1514
+ "step": 5375
1515
+ },
1516
+ {
1517
+ "epoch": 2.23,
1518
+ "grad_norm": 0.32012248039245605,
1519
+ "learning_rate": 1.4731281580156178e-05,
1520
+ "loss": 0.0447,
1521
+ "step": 5400
1522
+ },
1523
+ {
1524
+ "epoch": 2.24,
1525
+ "grad_norm": 0.3855270445346832,
1526
+ "learning_rate": 1.4645153881488286e-05,
1527
+ "loss": 0.0492,
1528
+ "step": 5425
1529
+ },
1530
+ {
1531
+ "epoch": 2.25,
1532
+ "grad_norm": 0.238910511136055,
1533
+ "learning_rate": 1.4559026182820396e-05,
1534
+ "loss": 0.0527,
1535
+ "step": 5450
1536
+ },
1537
+ {
1538
+ "epoch": 2.26,
1539
+ "grad_norm": 0.3347514569759369,
1540
+ "learning_rate": 1.4472898484152504e-05,
1541
+ "loss": 0.0514,
1542
+ "step": 5475
1543
+ },
1544
+ {
1545
+ "epoch": 2.27,
1546
+ "grad_norm": 0.34102463722229004,
1547
+ "learning_rate": 1.4386770785484612e-05,
1548
+ "loss": 0.05,
1549
+ "step": 5500
1550
+ },
1551
+ {
1552
+ "epoch": 2.28,
1553
+ "grad_norm": 0.4247712790966034,
1554
+ "learning_rate": 1.4304088194763437e-05,
1555
+ "loss": 0.0537,
1556
+ "step": 5525
1557
+ },
1558
+ {
1559
+ "epoch": 2.29,
1560
+ "grad_norm": 0.2721666395664215,
1561
+ "learning_rate": 1.4217960496095544e-05,
1562
+ "loss": 0.0503,
1563
+ "step": 5550
1564
+ },
1565
+ {
1566
+ "epoch": 2.3,
1567
+ "grad_norm": 0.4121835231781006,
1568
+ "learning_rate": 1.4131832797427654e-05,
1569
+ "loss": 0.0471,
1570
+ "step": 5575
1571
+ },
1572
+ {
1573
+ "epoch": 2.32,
1574
+ "grad_norm": 0.34142744541168213,
1575
+ "learning_rate": 1.4045705098759761e-05,
1576
+ "loss": 0.0528,
1577
+ "step": 5600
1578
+ },
1579
+ {
1580
+ "epoch": 2.33,
1581
+ "grad_norm": 0.44415149092674255,
1582
+ "learning_rate": 1.3959577400091869e-05,
1583
+ "loss": 0.0476,
1584
+ "step": 5625
1585
+ },
1586
+ {
1587
+ "epoch": 2.34,
1588
+ "grad_norm": 0.4206011891365051,
1589
+ "learning_rate": 1.3873449701423978e-05,
1590
+ "loss": 0.0476,
1591
+ "step": 5650
1592
+ },
1593
+ {
1594
+ "epoch": 2.35,
1595
+ "grad_norm": 0.2635906934738159,
1596
+ "learning_rate": 1.3787322002756087e-05,
1597
+ "loss": 0.0412,
1598
+ "step": 5675
1599
+ },
1600
+ {
1601
+ "epoch": 2.36,
1602
+ "grad_norm": 0.21374674141407013,
1603
+ "learning_rate": 1.3701194304088195e-05,
1604
+ "loss": 0.0476,
1605
+ "step": 5700
1606
+ },
1607
+ {
1608
+ "epoch": 2.37,
1609
+ "grad_norm": 0.29949426651000977,
1610
+ "learning_rate": 1.3615066605420304e-05,
1611
+ "loss": 0.0526,
1612
+ "step": 5725
1613
+ },
1614
+ {
1615
+ "epoch": 2.38,
1616
+ "grad_norm": 0.28353065252304077,
1617
+ "learning_rate": 1.3528938906752411e-05,
1618
+ "loss": 0.0523,
1619
+ "step": 5750
1620
+ },
1621
+ {
1622
+ "epoch": 2.39,
1623
+ "grad_norm": 0.37691453099250793,
1624
+ "learning_rate": 1.344281120808452e-05,
1625
+ "loss": 0.0496,
1626
+ "step": 5775
1627
+ },
1628
+ {
1629
+ "epoch": 2.4,
1630
+ "grad_norm": 0.20896878838539124,
1631
+ "learning_rate": 1.3356683509416628e-05,
1632
+ "loss": 0.0505,
1633
+ "step": 5800
1634
+ },
1635
+ {
1636
+ "epoch": 2.41,
1637
+ "grad_norm": 0.2891576588153839,
1638
+ "learning_rate": 1.3270555810748737e-05,
1639
+ "loss": 0.0519,
1640
+ "step": 5825
1641
+ },
1642
+ {
1643
+ "epoch": 2.42,
1644
+ "grad_norm": 0.32109466195106506,
1645
+ "learning_rate": 1.3184428112080847e-05,
1646
+ "loss": 0.0457,
1647
+ "step": 5850
1648
+ },
1649
+ {
1650
+ "epoch": 2.43,
1651
+ "grad_norm": 0.37973764538764954,
1652
+ "learning_rate": 1.3098300413412954e-05,
1653
+ "loss": 0.0469,
1654
+ "step": 5875
1655
+ },
1656
+ {
1657
+ "epoch": 2.44,
1658
+ "grad_norm": 0.31194964051246643,
1659
+ "learning_rate": 1.3012172714745062e-05,
1660
+ "loss": 0.0455,
1661
+ "step": 5900
1662
+ },
1663
+ {
1664
+ "epoch": 2.45,
1665
+ "grad_norm": 0.46967265009880066,
1666
+ "learning_rate": 1.2926045016077171e-05,
1667
+ "loss": 0.0503,
1668
+ "step": 5925
1669
+ },
1670
+ {
1671
+ "epoch": 2.46,
1672
+ "grad_norm": 0.23756754398345947,
1673
+ "learning_rate": 1.283991731740928e-05,
1674
+ "loss": 0.0534,
1675
+ "step": 5950
1676
+ },
1677
+ {
1678
+ "epoch": 2.47,
1679
+ "grad_norm": 0.2805767059326172,
1680
+ "learning_rate": 1.2753789618741388e-05,
1681
+ "loss": 0.0535,
1682
+ "step": 5975
1683
+ },
1684
+ {
1685
+ "epoch": 2.48,
1686
+ "grad_norm": 0.3886590301990509,
1687
+ "learning_rate": 1.2667661920073497e-05,
1688
+ "loss": 0.0483,
1689
+ "step": 6000
1690
+ },
1691
+ {
1692
+ "epoch": 2.49,
1693
+ "grad_norm": 0.35990557074546814,
1694
+ "learning_rate": 1.2581534221405604e-05,
1695
+ "loss": 0.0442,
1696
+ "step": 6025
1697
+ },
1698
+ {
1699
+ "epoch": 2.5,
1700
+ "grad_norm": 0.39960598945617676,
1701
+ "learning_rate": 1.2495406522737712e-05,
1702
+ "loss": 0.0517,
1703
+ "step": 6050
1704
+ },
1705
+ {
1706
+ "epoch": 2.51,
1707
+ "grad_norm": 0.25265923142433167,
1708
+ "learning_rate": 1.2409278824069821e-05,
1709
+ "loss": 0.052,
1710
+ "step": 6075
1711
+ },
1712
+ {
1713
+ "epoch": 2.52,
1714
+ "grad_norm": 0.19778184592723846,
1715
+ "learning_rate": 1.232315112540193e-05,
1716
+ "loss": 0.0489,
1717
+ "step": 6100
1718
+ },
1719
+ {
1720
+ "epoch": 2.53,
1721
+ "grad_norm": 0.3069133758544922,
1722
+ "learning_rate": 1.2237023426734038e-05,
1723
+ "loss": 0.05,
1724
+ "step": 6125
1725
+ },
1726
+ {
1727
+ "epoch": 2.54,
1728
+ "grad_norm": 0.23380334675312042,
1729
+ "learning_rate": 1.2150895728066147e-05,
1730
+ "loss": 0.0456,
1731
+ "step": 6150
1732
+ },
1733
+ {
1734
+ "epoch": 2.55,
1735
+ "grad_norm": 0.22880606353282928,
1736
+ "learning_rate": 1.2064768029398255e-05,
1737
+ "loss": 0.0567,
1738
+ "step": 6175
1739
+ },
1740
+ {
1741
+ "epoch": 2.56,
1742
+ "grad_norm": 0.2991226315498352,
1743
+ "learning_rate": 1.1978640330730362e-05,
1744
+ "loss": 0.0537,
1745
+ "step": 6200
1746
+ },
1747
+ {
1748
+ "epoch": 2.57,
1749
+ "grad_norm": 0.23710553348064423,
1750
+ "learning_rate": 1.1892512632062471e-05,
1751
+ "loss": 0.0453,
1752
+ "step": 6225
1753
+ },
1754
+ {
1755
+ "epoch": 2.58,
1756
+ "grad_norm": 0.37337175011634827,
1757
+ "learning_rate": 1.180638493339458e-05,
1758
+ "loss": 0.0554,
1759
+ "step": 6250
1760
+ },
1761
+ {
1762
+ "epoch": 2.59,
1763
+ "grad_norm": 0.2040768265724182,
1764
+ "learning_rate": 1.1720257234726688e-05,
1765
+ "loss": 0.0568,
1766
+ "step": 6275
1767
+ },
1768
+ {
1769
+ "epoch": 2.6,
1770
+ "grad_norm": 0.14503860473632812,
1771
+ "learning_rate": 1.1634129536058797e-05,
1772
+ "loss": 0.0488,
1773
+ "step": 6300
1774
+ },
1775
+ {
1776
+ "epoch": 2.61,
1777
+ "grad_norm": 0.41582000255584717,
1778
+ "learning_rate": 1.1548001837390905e-05,
1779
+ "loss": 0.0503,
1780
+ "step": 6325
1781
+ },
1782
+ {
1783
+ "epoch": 2.63,
1784
+ "grad_norm": 0.4040040075778961,
1785
+ "learning_rate": 1.1461874138723012e-05,
1786
+ "loss": 0.0448,
1787
+ "step": 6350
1788
+ },
1789
+ {
1790
+ "epoch": 2.64,
1791
+ "grad_norm": 0.37314775586128235,
1792
+ "learning_rate": 1.1375746440055123e-05,
1793
+ "loss": 0.051,
1794
+ "step": 6375
1795
+ },
1796
+ {
1797
+ "epoch": 2.65,
1798
+ "grad_norm": 0.37785276770591736,
1799
+ "learning_rate": 1.128961874138723e-05,
1800
+ "loss": 0.0524,
1801
+ "step": 6400
1802
+ },
1803
+ {
1804
+ "epoch": 2.66,
1805
+ "grad_norm": 0.12397664785385132,
1806
+ "learning_rate": 1.1203491042719338e-05,
1807
+ "loss": 0.0554,
1808
+ "step": 6425
1809
+ },
1810
+ {
1811
+ "epoch": 2.67,
1812
+ "grad_norm": 0.252644419670105,
1813
+ "learning_rate": 1.1117363344051448e-05,
1814
+ "loss": 0.0504,
1815
+ "step": 6450
1816
+ },
1817
+ {
1818
+ "epoch": 2.68,
1819
+ "grad_norm": 0.27793678641319275,
1820
+ "learning_rate": 1.1031235645383555e-05,
1821
+ "loss": 0.0484,
1822
+ "step": 6475
1823
+ },
1824
+ {
1825
+ "epoch": 2.69,
1826
+ "grad_norm": 0.27731946110725403,
1827
+ "learning_rate": 1.0945107946715663e-05,
1828
+ "loss": 0.0456,
1829
+ "step": 6500
1830
+ },
1831
+ {
1832
+ "epoch": 2.7,
1833
+ "grad_norm": 0.5174788236618042,
1834
+ "learning_rate": 1.0858980248047774e-05,
1835
+ "loss": 0.0514,
1836
+ "step": 6525
1837
+ },
1838
+ {
1839
+ "epoch": 2.71,
1840
+ "grad_norm": 0.3238148093223572,
1841
+ "learning_rate": 1.0772852549379881e-05,
1842
+ "loss": 0.0509,
1843
+ "step": 6550
1844
+ },
1845
+ {
1846
+ "epoch": 2.72,
1847
+ "grad_norm": 0.25130346417427063,
1848
+ "learning_rate": 1.0686724850711989e-05,
1849
+ "loss": 0.046,
1850
+ "step": 6575
1851
+ },
1852
+ {
1853
+ "epoch": 2.73,
1854
+ "grad_norm": 0.2648930549621582,
1855
+ "learning_rate": 1.0600597152044098e-05,
1856
+ "loss": 0.0482,
1857
+ "step": 6600
1858
+ },
1859
+ {
1860
+ "epoch": 2.74,
1861
+ "grad_norm": 0.23606421053409576,
1862
+ "learning_rate": 1.0514469453376205e-05,
1863
+ "loss": 0.049,
1864
+ "step": 6625
1865
+ },
1866
+ {
1867
+ "epoch": 2.75,
1868
+ "grad_norm": 0.13270416855812073,
1869
+ "learning_rate": 1.0428341754708315e-05,
1870
+ "loss": 0.0494,
1871
+ "step": 6650
1872
+ },
1873
+ {
1874
+ "epoch": 2.76,
1875
+ "grad_norm": 0.485784113407135,
1876
+ "learning_rate": 1.0342214056040424e-05,
1877
+ "loss": 0.0435,
1878
+ "step": 6675
1879
+ },
1880
+ {
1881
+ "epoch": 2.77,
1882
+ "grad_norm": 0.37061092257499695,
1883
+ "learning_rate": 1.0256086357372531e-05,
1884
+ "loss": 0.0458,
1885
+ "step": 6700
1886
+ },
1887
+ {
1888
+ "epoch": 2.78,
1889
+ "grad_norm": 0.23628602921962738,
1890
+ "learning_rate": 1.0169958658704639e-05,
1891
+ "loss": 0.0488,
1892
+ "step": 6725
1893
+ },
1894
+ {
1895
+ "epoch": 2.79,
1896
+ "grad_norm": 0.5137524604797363,
1897
+ "learning_rate": 1.0083830960036748e-05,
1898
+ "loss": 0.0515,
1899
+ "step": 6750
1900
+ },
1901
+ {
1902
+ "epoch": 2.8,
1903
+ "grad_norm": 0.4176236093044281,
1904
+ "learning_rate": 9.997703261368856e-06,
1905
+ "loss": 0.0452,
1906
+ "step": 6775
1907
+ },
1908
+ {
1909
+ "epoch": 2.81,
1910
+ "grad_norm": 0.3956039547920227,
1911
+ "learning_rate": 9.911575562700965e-06,
1912
+ "loss": 0.0499,
1913
+ "step": 6800
1914
+ },
1915
+ {
1916
+ "epoch": 2.82,
1917
+ "grad_norm": 0.29062360525131226,
1918
+ "learning_rate": 9.825447864033074e-06,
1919
+ "loss": 0.0479,
1920
+ "step": 6825
1921
+ },
1922
+ {
1923
+ "epoch": 2.83,
1924
+ "grad_norm": 0.3083875775337219,
1925
+ "learning_rate": 9.739320165365182e-06,
1926
+ "loss": 0.0501,
1927
+ "step": 6850
1928
+ },
1929
+ {
1930
+ "epoch": 2.84,
1931
+ "grad_norm": 0.43474555015563965,
1932
+ "learning_rate": 9.653192466697289e-06,
1933
+ "loss": 0.0562,
1934
+ "step": 6875
1935
+ },
1936
+ {
1937
+ "epoch": 2.85,
1938
+ "grad_norm": 0.21572205424308777,
1939
+ "learning_rate": 9.567064768029398e-06,
1940
+ "loss": 0.049,
1941
+ "step": 6900
1942
+ },
1943
+ {
1944
+ "epoch": 2.86,
1945
+ "grad_norm": 0.17227627336978912,
1946
+ "learning_rate": 9.480937069361506e-06,
1947
+ "loss": 0.0455,
1948
+ "step": 6925
1949
+ },
1950
+ {
1951
+ "epoch": 2.87,
1952
+ "grad_norm": 0.34257158637046814,
1953
+ "learning_rate": 9.394809370693617e-06,
1954
+ "loss": 0.0456,
1955
+ "step": 6950
1956
+ },
1957
+ {
1958
+ "epoch": 2.88,
1959
+ "grad_norm": 0.30878308415412903,
1960
+ "learning_rate": 9.308681672025724e-06,
1961
+ "loss": 0.0469,
1962
+ "step": 6975
1963
+ },
1964
+ {
1965
+ "epoch": 2.89,
1966
+ "grad_norm": 0.338555246591568,
1967
+ "learning_rate": 9.222553973357832e-06,
1968
+ "loss": 0.0458,
1969
+ "step": 7000
1970
+ },
1971
+ {
1972
+ "epoch": 2.9,
1973
+ "grad_norm": 0.2819863259792328,
1974
+ "learning_rate": 9.136426274689941e-06,
1975
+ "loss": 0.0514,
1976
+ "step": 7025
1977
+ },
1978
+ {
1979
+ "epoch": 2.91,
1980
+ "grad_norm": 0.27121391892433167,
1981
+ "learning_rate": 9.050298576022049e-06,
1982
+ "loss": 0.0574,
1983
+ "step": 7050
1984
+ },
1985
+ {
1986
+ "epoch": 2.92,
1987
+ "grad_norm": 0.28231513500213623,
1988
+ "learning_rate": 8.964170877354158e-06,
1989
+ "loss": 0.0459,
1990
+ "step": 7075
1991
+ },
1992
+ {
1993
+ "epoch": 2.94,
1994
+ "grad_norm": 0.3147279620170593,
1995
+ "learning_rate": 8.878043178686267e-06,
1996
+ "loss": 0.055,
1997
+ "step": 7100
1998
+ },
1999
+ {
2000
+ "epoch": 2.95,
2001
+ "grad_norm": 0.3944794535636902,
2002
+ "learning_rate": 8.791915480018375e-06,
2003
+ "loss": 0.0561,
2004
+ "step": 7125
2005
+ },
2006
+ {
2007
+ "epoch": 2.96,
2008
+ "grad_norm": 0.5070275664329529,
2009
+ "learning_rate": 8.705787781350482e-06,
2010
+ "loss": 0.0452,
2011
+ "step": 7150
2012
+ },
2013
+ {
2014
+ "epoch": 2.97,
2015
+ "grad_norm": 0.29337751865386963,
2016
+ "learning_rate": 8.619660082682591e-06,
2017
+ "loss": 0.047,
2018
+ "step": 7175
2019
+ },
2020
+ {
2021
+ "epoch": 2.98,
2022
+ "grad_norm": 0.3418371081352234,
2023
+ "learning_rate": 8.533532384014699e-06,
2024
+ "loss": 0.0441,
2025
+ "step": 7200
2026
+ },
2027
+ {
2028
+ "epoch": 2.99,
2029
+ "grad_norm": 0.34925562143325806,
2030
+ "learning_rate": 8.447404685346808e-06,
2031
+ "loss": 0.043,
2032
+ "step": 7225
2033
+ },
2034
+ {
2035
+ "epoch": 3.0,
2036
+ "grad_norm": 0.32116442918777466,
2037
+ "learning_rate": 8.361276986678917e-06,
2038
+ "loss": 0.0475,
2039
+ "step": 7250
2040
+ },
2041
+ {
2042
+ "epoch": 3.01,
2043
+ "grad_norm": 0.3370840549468994,
2044
+ "learning_rate": 8.275149288011025e-06,
2045
+ "loss": 0.0454,
2046
+ "step": 7275
2047
+ },
2048
+ {
2049
+ "epoch": 3.02,
2050
+ "grad_norm": 0.28512734174728394,
2051
+ "learning_rate": 8.189021589343132e-06,
2052
+ "loss": 0.0358,
2053
+ "step": 7300
2054
+ },
2055
+ {
2056
+ "epoch": 3.03,
2057
+ "grad_norm": 0.35008323192596436,
2058
+ "learning_rate": 8.102893890675242e-06,
2059
+ "loss": 0.0425,
2060
+ "step": 7325
2061
+ },
2062
+ {
2063
+ "epoch": 3.04,
2064
+ "grad_norm": 0.32068151235580444,
2065
+ "learning_rate": 8.016766192007349e-06,
2066
+ "loss": 0.0413,
2067
+ "step": 7350
2068
+ },
2069
+ {
2070
+ "epoch": 3.05,
2071
+ "grad_norm": 0.27377787232398987,
2072
+ "learning_rate": 7.930638493339458e-06,
2073
+ "loss": 0.037,
2074
+ "step": 7375
2075
+ },
2076
+ {
2077
+ "epoch": 3.06,
2078
+ "grad_norm": 0.1453736424446106,
2079
+ "learning_rate": 7.844510794671568e-06,
2080
+ "loss": 0.0364,
2081
+ "step": 7400
2082
+ },
2083
+ {
2084
+ "epoch": 3.07,
2085
+ "grad_norm": 0.4593455493450165,
2086
+ "learning_rate": 7.758383096003675e-06,
2087
+ "loss": 0.0398,
2088
+ "step": 7425
2089
+ },
2090
+ {
2091
+ "epoch": 3.08,
2092
+ "grad_norm": 0.3376910984516144,
2093
+ "learning_rate": 7.672255397335783e-06,
2094
+ "loss": 0.0425,
2095
+ "step": 7450
2096
+ },
2097
+ {
2098
+ "epoch": 3.09,
2099
+ "grad_norm": 0.4269729554653168,
2100
+ "learning_rate": 7.586127698667893e-06,
2101
+ "loss": 0.0421,
2102
+ "step": 7475
2103
+ },
2104
+ {
2105
+ "epoch": 3.1,
2106
+ "grad_norm": 0.42669641971588135,
2107
+ "learning_rate": 7.5e-06,
2108
+ "loss": 0.0402,
2109
+ "step": 7500
2110
+ },
2111
+ {
2112
+ "epoch": 3.11,
2113
+ "grad_norm": 0.3629492223262787,
2114
+ "learning_rate": 7.4138723013321086e-06,
2115
+ "loss": 0.0379,
2116
+ "step": 7525
2117
+ },
2118
+ {
2119
+ "epoch": 3.12,
2120
+ "grad_norm": 0.24746298789978027,
2121
+ "learning_rate": 7.327744602664217e-06,
2122
+ "loss": 0.0404,
2123
+ "step": 7550
2124
+ },
2125
+ {
2126
+ "epoch": 3.13,
2127
+ "grad_norm": 0.19069620966911316,
2128
+ "learning_rate": 7.241616903996325e-06,
2129
+ "loss": 0.0393,
2130
+ "step": 7575
2131
+ },
2132
+ {
2133
+ "epoch": 3.14,
2134
+ "grad_norm": 0.2659347951412201,
2135
+ "learning_rate": 7.155489205328434e-06,
2136
+ "loss": 0.0404,
2137
+ "step": 7600
2138
+ },
2139
+ {
2140
+ "epoch": 3.15,
2141
+ "grad_norm": 0.15745492279529572,
2142
+ "learning_rate": 7.069361506660542e-06,
2143
+ "loss": 0.0392,
2144
+ "step": 7625
2145
+ },
2146
+ {
2147
+ "epoch": 3.16,
2148
+ "grad_norm": 0.20051732659339905,
2149
+ "learning_rate": 6.9866789159393665e-06,
2150
+ "loss": 0.0361,
2151
+ "step": 7650
2152
+ },
2153
+ {
2154
+ "epoch": 3.17,
2155
+ "grad_norm": 0.3713512122631073,
2156
+ "learning_rate": 6.900551217271474e-06,
2157
+ "loss": 0.0386,
2158
+ "step": 7675
2159
+ },
2160
+ {
2161
+ "epoch": 3.18,
2162
+ "grad_norm": 0.42229992151260376,
2163
+ "learning_rate": 6.814423518603583e-06,
2164
+ "loss": 0.0361,
2165
+ "step": 7700
2166
+ },
2167
+ {
2168
+ "epoch": 3.19,
2169
+ "grad_norm": 0.2441110610961914,
2170
+ "learning_rate": 6.728295819935692e-06,
2171
+ "loss": 0.037,
2172
+ "step": 7725
2173
+ },
2174
+ {
2175
+ "epoch": 3.2,
2176
+ "grad_norm": 0.18802495300769806,
2177
+ "learning_rate": 6.642168121267799e-06,
2178
+ "loss": 0.042,
2179
+ "step": 7750
2180
+ },
2181
+ {
2182
+ "epoch": 3.21,
2183
+ "grad_norm": 0.25679391622543335,
2184
+ "learning_rate": 6.556040422599908e-06,
2185
+ "loss": 0.0393,
2186
+ "step": 7775
2187
+ },
2188
+ {
2189
+ "epoch": 3.22,
2190
+ "grad_norm": 0.29285240173339844,
2191
+ "learning_rate": 6.469912723932017e-06,
2192
+ "loss": 0.0419,
2193
+ "step": 7800
2194
+ },
2195
+ {
2196
+ "epoch": 3.23,
2197
+ "grad_norm": 0.3169197738170624,
2198
+ "learning_rate": 6.383785025264124e-06,
2199
+ "loss": 0.0409,
2200
+ "step": 7825
2201
+ },
2202
+ {
2203
+ "epoch": 3.25,
2204
+ "grad_norm": 0.32817521691322327,
2205
+ "learning_rate": 6.2976573265962335e-06,
2206
+ "loss": 0.0403,
2207
+ "step": 7850
2208
+ },
2209
+ {
2210
+ "epoch": 3.26,
2211
+ "grad_norm": 0.29750293493270874,
2212
+ "learning_rate": 6.211529627928342e-06,
2213
+ "loss": 0.0387,
2214
+ "step": 7875
2215
+ },
2216
+ {
2217
+ "epoch": 3.27,
2218
+ "grad_norm": 0.44555339217185974,
2219
+ "learning_rate": 6.125401929260451e-06,
2220
+ "loss": 0.0413,
2221
+ "step": 7900
2222
+ },
2223
+ {
2224
+ "epoch": 3.28,
2225
+ "grad_norm": 0.34628069400787354,
2226
+ "learning_rate": 6.039274230592559e-06,
2227
+ "loss": 0.0405,
2228
+ "step": 7925
2229
+ },
2230
+ {
2231
+ "epoch": 3.29,
2232
+ "grad_norm": 0.36807015538215637,
2233
+ "learning_rate": 5.953146531924667e-06,
2234
+ "loss": 0.0399,
2235
+ "step": 7950
2236
+ },
2237
+ {
2238
+ "epoch": 3.3,
2239
+ "grad_norm": 0.2611570656299591,
2240
+ "learning_rate": 5.867018833256776e-06,
2241
+ "loss": 0.0384,
2242
+ "step": 7975
2243
+ },
2244
+ {
2245
+ "epoch": 3.31,
2246
+ "grad_norm": 0.2191769927740097,
2247
+ "learning_rate": 5.780891134588884e-06,
2248
+ "loss": 0.0387,
2249
+ "step": 8000
2250
+ },
2251
+ {
2252
+ "epoch": 3.32,
2253
+ "grad_norm": 0.38859057426452637,
2254
+ "learning_rate": 5.694763435920992e-06,
2255
+ "loss": 0.0451,
2256
+ "step": 8025
2257
+ },
2258
+ {
2259
+ "epoch": 3.33,
2260
+ "grad_norm": 0.24283719062805176,
2261
+ "learning_rate": 5.608635737253101e-06,
2262
+ "loss": 0.0378,
2263
+ "step": 8050
2264
+ },
2265
+ {
2266
+ "epoch": 3.34,
2267
+ "grad_norm": 0.3681369423866272,
2268
+ "learning_rate": 5.522508038585209e-06,
2269
+ "loss": 0.0438,
2270
+ "step": 8075
2271
+ },
2272
+ {
2273
+ "epoch": 3.35,
2274
+ "grad_norm": 0.3467373251914978,
2275
+ "learning_rate": 5.436380339917317e-06,
2276
+ "loss": 0.0416,
2277
+ "step": 8100
2278
+ },
2279
+ {
2280
+ "epoch": 3.36,
2281
+ "grad_norm": 0.3951582610607147,
2282
+ "learning_rate": 5.3502526412494265e-06,
2283
+ "loss": 0.0407,
2284
+ "step": 8125
2285
+ },
2286
+ {
2287
+ "epoch": 3.37,
2288
+ "grad_norm": 0.581302285194397,
2289
+ "learning_rate": 5.264124942581534e-06,
2290
+ "loss": 0.0391,
2291
+ "step": 8150
2292
+ },
2293
+ {
2294
+ "epoch": 3.38,
2295
+ "grad_norm": 0.34989818930625916,
2296
+ "learning_rate": 5.177997243913642e-06,
2297
+ "loss": 0.0415,
2298
+ "step": 8175
2299
+ },
2300
+ {
2301
+ "epoch": 3.39,
2302
+ "grad_norm": 0.4126095771789551,
2303
+ "learning_rate": 5.091869545245752e-06,
2304
+ "loss": 0.0364,
2305
+ "step": 8200
2306
+ },
2307
+ {
2308
+ "epoch": 3.4,
2309
+ "grad_norm": 0.3435223698616028,
2310
+ "learning_rate": 5.005741846577859e-06,
2311
+ "loss": 0.0375,
2312
+ "step": 8225
2313
+ },
2314
+ {
2315
+ "epoch": 3.41,
2316
+ "grad_norm": 0.42563363909721375,
2317
+ "learning_rate": 4.9196141479099675e-06,
2318
+ "loss": 0.0404,
2319
+ "step": 8250
2320
+ },
2321
+ {
2322
+ "epoch": 3.42,
2323
+ "grad_norm": 0.2580004632472992,
2324
+ "learning_rate": 4.833486449242077e-06,
2325
+ "loss": 0.0381,
2326
+ "step": 8275
2327
+ },
2328
+ {
2329
+ "epoch": 3.43,
2330
+ "grad_norm": 0.37050294876098633,
2331
+ "learning_rate": 4.747358750574184e-06,
2332
+ "loss": 0.042,
2333
+ "step": 8300
2334
+ },
2335
+ {
2336
+ "epoch": 3.44,
2337
+ "grad_norm": 0.31261393427848816,
2338
+ "learning_rate": 4.6612310519062935e-06,
2339
+ "loss": 0.0411,
2340
+ "step": 8325
2341
+ },
2342
+ {
2343
+ "epoch": 3.45,
2344
+ "grad_norm": 0.33394724130630493,
2345
+ "learning_rate": 4.575103353238402e-06,
2346
+ "loss": 0.0445,
2347
+ "step": 8350
2348
+ },
2349
+ {
2350
+ "epoch": 3.46,
2351
+ "grad_norm": 0.6875510811805725,
2352
+ "learning_rate": 4.4889756545705094e-06,
2353
+ "loss": 0.0407,
2354
+ "step": 8375
2355
+ },
2356
+ {
2357
+ "epoch": 3.47,
2358
+ "grad_norm": 0.4609951674938202,
2359
+ "learning_rate": 4.402847955902619e-06,
2360
+ "loss": 0.0355,
2361
+ "step": 8400
2362
+ },
2363
+ {
2364
+ "epoch": 3.48,
2365
+ "grad_norm": 0.36755943298339844,
2366
+ "learning_rate": 4.316720257234727e-06,
2367
+ "loss": 0.0412,
2368
+ "step": 8425
2369
+ },
2370
+ {
2371
+ "epoch": 3.49,
2372
+ "grad_norm": 0.3488776385784149,
2373
+ "learning_rate": 4.2305925585668345e-06,
2374
+ "loss": 0.0432,
2375
+ "step": 8450
2376
+ },
2377
+ {
2378
+ "epoch": 3.5,
2379
+ "grad_norm": 0.2769566476345062,
2380
+ "learning_rate": 4.144464859898944e-06,
2381
+ "loss": 0.034,
2382
+ "step": 8475
2383
+ },
2384
+ {
2385
+ "epoch": 3.51,
2386
+ "grad_norm": 0.3229292333126068,
2387
+ "learning_rate": 4.058337161231052e-06,
2388
+ "loss": 0.0351,
2389
+ "step": 8500
2390
+ },
2391
+ {
2392
+ "epoch": 3.52,
2393
+ "grad_norm": 0.39641430974006653,
2394
+ "learning_rate": 3.9722094625631605e-06,
2395
+ "loss": 0.0365,
2396
+ "step": 8525
2397
+ },
2398
+ {
2399
+ "epoch": 3.53,
2400
+ "grad_norm": 0.3574431240558624,
2401
+ "learning_rate": 3.886081763895269e-06,
2402
+ "loss": 0.04,
2403
+ "step": 8550
2404
+ },
2405
+ {
2406
+ "epoch": 3.54,
2407
+ "grad_norm": 0.3420943319797516,
2408
+ "learning_rate": 3.7999540652273773e-06,
2409
+ "loss": 0.039,
2410
+ "step": 8575
2411
+ },
2412
+ {
2413
+ "epoch": 3.56,
2414
+ "grad_norm": 0.3045111894607544,
2415
+ "learning_rate": 3.7138263665594856e-06,
2416
+ "loss": 0.0394,
2417
+ "step": 8600
2418
+ },
2419
+ {
2420
+ "epoch": 3.57,
2421
+ "grad_norm": 0.5104286670684814,
2422
+ "learning_rate": 3.627698667891594e-06,
2423
+ "loss": 0.0358,
2424
+ "step": 8625
2425
+ },
2426
+ {
2427
+ "epoch": 3.58,
2428
+ "grad_norm": 0.2855064272880554,
2429
+ "learning_rate": 3.5415709692237024e-06,
2430
+ "loss": 0.0426,
2431
+ "step": 8650
2432
+ },
2433
+ {
2434
+ "epoch": 3.59,
2435
+ "grad_norm": 0.3009551167488098,
2436
+ "learning_rate": 3.4554432705558108e-06,
2437
+ "loss": 0.0412,
2438
+ "step": 8675
2439
+ },
2440
+ {
2441
+ "epoch": 3.6,
2442
+ "grad_norm": 0.24823172390460968,
2443
+ "learning_rate": 3.3693155718879196e-06,
2444
+ "loss": 0.0408,
2445
+ "step": 8700
2446
+ },
2447
+ {
2448
+ "epoch": 3.61,
2449
+ "grad_norm": 0.3101685643196106,
2450
+ "learning_rate": 3.2831878732200275e-06,
2451
+ "loss": 0.0385,
2452
+ "step": 8725
2453
+ },
2454
+ {
2455
+ "epoch": 3.62,
2456
+ "grad_norm": 0.2525741159915924,
2457
+ "learning_rate": 3.197060174552136e-06,
2458
+ "loss": 0.0405,
2459
+ "step": 8750
2460
+ },
2461
+ {
2462
+ "epoch": 3.63,
2463
+ "grad_norm": 0.27567043900489807,
2464
+ "learning_rate": 3.1109324758842447e-06,
2465
+ "loss": 0.0402,
2466
+ "step": 8775
2467
+ },
2468
+ {
2469
+ "epoch": 3.64,
2470
+ "grad_norm": 0.3665957748889923,
2471
+ "learning_rate": 3.0248047772163526e-06,
2472
+ "loss": 0.0369,
2473
+ "step": 8800
2474
+ },
2475
+ {
2476
+ "epoch": 3.65,
2477
+ "grad_norm": 0.32239583134651184,
2478
+ "learning_rate": 2.938677078548461e-06,
2479
+ "loss": 0.0366,
2480
+ "step": 8825
2481
+ },
2482
+ {
2483
+ "epoch": 3.66,
2484
+ "grad_norm": 0.2736150026321411,
2485
+ "learning_rate": 2.85254937988057e-06,
2486
+ "loss": 0.0379,
2487
+ "step": 8850
2488
+ },
2489
+ {
2490
+ "epoch": 3.67,
2491
+ "grad_norm": 0.6374419331550598,
2492
+ "learning_rate": 2.766421681212678e-06,
2493
+ "loss": 0.0386,
2494
+ "step": 8875
2495
+ },
2496
+ {
2497
+ "epoch": 3.68,
2498
+ "grad_norm": 0.17933352291584015,
2499
+ "learning_rate": 2.680293982544786e-06,
2500
+ "loss": 0.0403,
2501
+ "step": 8900
2502
+ },
2503
+ {
2504
+ "epoch": 3.69,
2505
+ "grad_norm": 0.429572731256485,
2506
+ "learning_rate": 2.594166283876895e-06,
2507
+ "loss": 0.041,
2508
+ "step": 8925
2509
+ },
2510
+ {
2511
+ "epoch": 3.7,
2512
+ "grad_norm": 0.343364953994751,
2513
+ "learning_rate": 2.5080385852090033e-06,
2514
+ "loss": 0.0415,
2515
+ "step": 8950
2516
+ },
2517
+ {
2518
+ "epoch": 3.71,
2519
+ "grad_norm": 0.3911416530609131,
2520
+ "learning_rate": 2.4219108865411117e-06,
2521
+ "loss": 0.0337,
2522
+ "step": 8975
2523
+ },
2524
+ {
2525
+ "epoch": 3.72,
2526
+ "grad_norm": 0.43486475944519043,
2527
+ "learning_rate": 2.33578318787322e-06,
2528
+ "loss": 0.0415,
2529
+ "step": 9000
2530
+ },
2531
+ {
2532
+ "epoch": 3.73,
2533
+ "grad_norm": 0.22443470358848572,
2534
+ "learning_rate": 2.2496554892053284e-06,
2535
+ "loss": 0.0389,
2536
+ "step": 9025
2537
+ },
2538
+ {
2539
+ "epoch": 3.74,
2540
+ "grad_norm": 0.36421477794647217,
2541
+ "learning_rate": 2.1635277905374372e-06,
2542
+ "loss": 0.0407,
2543
+ "step": 9050
2544
+ },
2545
+ {
2546
+ "epoch": 3.75,
2547
+ "grad_norm": 0.382355660200119,
2548
+ "learning_rate": 2.077400091869545e-06,
2549
+ "loss": 0.0394,
2550
+ "step": 9075
2551
+ },
2552
+ {
2553
+ "epoch": 3.76,
2554
+ "grad_norm": 0.2957639694213867,
2555
+ "learning_rate": 1.9912723932016536e-06,
2556
+ "loss": 0.0425,
2557
+ "step": 9100
2558
+ },
2559
+ {
2560
+ "epoch": 3.77,
2561
+ "grad_norm": 0.44382327795028687,
2562
+ "learning_rate": 1.9051446945337622e-06,
2563
+ "loss": 0.0392,
2564
+ "step": 9125
2565
+ },
2566
+ {
2567
+ "epoch": 3.78,
2568
+ "grad_norm": 0.3849116265773773,
2569
+ "learning_rate": 1.8190169958658705e-06,
2570
+ "loss": 0.0376,
2571
+ "step": 9150
2572
+ },
2573
+ {
2574
+ "epoch": 3.79,
2575
+ "grad_norm": 0.37635913491249084,
2576
+ "learning_rate": 1.732889297197979e-06,
2577
+ "loss": 0.0364,
2578
+ "step": 9175
2579
+ },
2580
+ {
2581
+ "epoch": 3.8,
2582
+ "grad_norm": 0.2440154105424881,
2583
+ "learning_rate": 1.6467615985300875e-06,
2584
+ "loss": 0.0353,
2585
+ "step": 9200
2586
+ },
2587
+ {
2588
+ "epoch": 3.81,
2589
+ "grad_norm": 0.3721398711204529,
2590
+ "learning_rate": 1.5606338998621957e-06,
2591
+ "loss": 0.0456,
2592
+ "step": 9225
2593
+ },
2594
+ {
2595
+ "epoch": 3.82,
2596
+ "grad_norm": 0.42328134179115295,
2597
+ "learning_rate": 1.474506201194304e-06,
2598
+ "loss": 0.0381,
2599
+ "step": 9250
2600
+ },
2601
+ {
2602
+ "epoch": 3.83,
2603
+ "grad_norm": 0.3101789653301239,
2604
+ "learning_rate": 1.3883785025264126e-06,
2605
+ "loss": 0.0355,
2606
+ "step": 9275
2607
+ },
2608
+ {
2609
+ "epoch": 3.84,
2610
+ "grad_norm": 0.41222986578941345,
2611
+ "learning_rate": 1.302250803858521e-06,
2612
+ "loss": 0.0379,
2613
+ "step": 9300
2614
+ },
2615
+ {
2616
+ "epoch": 3.85,
2617
+ "grad_norm": 0.38471460342407227,
2618
+ "learning_rate": 1.2161231051906294e-06,
2619
+ "loss": 0.04,
2620
+ "step": 9325
2621
+ },
2622
+ {
2623
+ "epoch": 3.87,
2624
+ "grad_norm": 0.3404221534729004,
2625
+ "learning_rate": 1.1299954065227377e-06,
2626
+ "loss": 0.0429,
2627
+ "step": 9350
2628
+ },
2629
+ {
2630
+ "epoch": 3.88,
2631
+ "grad_norm": 0.3508870303630829,
2632
+ "learning_rate": 1.0438677078548461e-06,
2633
+ "loss": 0.0422,
2634
+ "step": 9375
2635
+ },
2636
+ {
2637
+ "epoch": 3.89,
2638
+ "grad_norm": 0.3830946385860443,
2639
+ "learning_rate": 9.577400091869545e-07,
2640
+ "loss": 0.0354,
2641
+ "step": 9400
2642
+ },
2643
+ {
2644
+ "epoch": 3.9,
2645
+ "grad_norm": 0.2902699410915375,
2646
+ "learning_rate": 8.71612310519063e-07,
2647
+ "loss": 0.0383,
2648
+ "step": 9425
2649
+ },
2650
+ {
2651
+ "epoch": 3.91,
2652
+ "grad_norm": 0.6969540119171143,
2653
+ "learning_rate": 7.854846118511714e-07,
2654
+ "loss": 0.0388,
2655
+ "step": 9450
2656
+ },
2657
+ {
2658
+ "epoch": 3.92,
2659
+ "grad_norm": 0.23226848244667053,
2660
+ "learning_rate": 6.993569131832797e-07,
2661
+ "loss": 0.0399,
2662
+ "step": 9475
2663
+ },
2664
+ {
2665
+ "epoch": 3.93,
2666
+ "grad_norm": 0.22804780304431915,
2667
+ "learning_rate": 6.132292145153882e-07,
2668
+ "loss": 0.0403,
2669
+ "step": 9500
2670
+ },
2671
+ {
2672
+ "epoch": 3.94,
2673
+ "grad_norm": 0.46192172169685364,
2674
+ "learning_rate": 5.271015158474966e-07,
2675
+ "loss": 0.0385,
2676
+ "step": 9525
2677
+ },
2678
+ {
2679
+ "epoch": 3.95,
2680
+ "grad_norm": 0.24761895835399628,
2681
+ "learning_rate": 4.4097381717960496e-07,
2682
+ "loss": 0.0372,
2683
+ "step": 9550
2684
+ },
2685
+ {
2686
+ "epoch": 3.96,
2687
+ "grad_norm": 0.3965108096599579,
2688
+ "learning_rate": 3.548461185117134e-07,
2689
+ "loss": 0.0389,
2690
+ "step": 9575
2691
+ },
2692
+ {
2693
+ "epoch": 3.97,
2694
+ "grad_norm": 0.24169424176216125,
2695
+ "learning_rate": 2.687184198438218e-07,
2696
+ "loss": 0.0393,
2697
+ "step": 9600
2698
+ },
2699
+ {
2700
+ "epoch": 3.98,
2701
+ "grad_norm": 0.4271424412727356,
2702
+ "learning_rate": 1.825907211759302e-07,
2703
+ "loss": 0.0385,
2704
+ "step": 9625
2705
+ },
2706
+ {
2707
+ "epoch": 3.99,
2708
+ "grad_norm": 0.3440055847167969,
2709
+ "learning_rate": 9.646302250803859e-08,
2710
+ "loss": 0.0394,
2711
+ "step": 9650
2712
+ },
2713
+ {
2714
+ "epoch": 4.0,
2715
+ "grad_norm": 0.43433114886283875,
2716
+ "learning_rate": 1.0335323840146991e-08,
2717
+ "loss": 0.036,
2718
+ "step": 9675
2719
+ }
2720
+ ],
2721
+ "logging_steps": 25,
2722
+ "max_steps": 9676,
2723
+ "num_input_tokens_seen": 0,
2724
+ "num_train_epochs": 4,
2725
+ "save_steps": 500,
2726
+ "total_flos": 8.653960459352801e+17,
2727
+ "train_batch_size": 2,
2728
+ "trial_name": null,
2729
+ "trial_params": null
2730
+ }
checkpoint-9676/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:471278576427115d6d17bfb25bab0702b85641a0b161bfd1e29e51fbba4b4223
3
+ size 4984
handler.py ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from typing import Dict, List, Any
2
+ from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
3
+ import torch
4
+ from peft import PeftModel
5
+ import json
6
+ import os
7
+
8
+
9
+ class EndpointHandler():
10
+ def __init__(self, path=""):
11
+ base_model_path = json.load(open(os.path.join(path, "training_params.json")))["model"]
12
+ model = AutoModelForCausalLM.from_pretrained(
13
+ base_model_path,
14
+ torch_dtype=torch.float16,
15
+ low_cpu_mem_usage=True,
16
+ trust_remote_code=True,
17
+ device_map="auto",
18
+ )
19
+ tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True)
20
+ model.resize_token_embeddings(len(tokenizer))
21
+ model = PeftModel.from_pretrained(model, path)
22
+ model = model.merge_and_unload()
23
+ self.pipeline = pipeline("text-generation", model=model, tokenizer=tokenizer)
24
+
25
+ def __call__(self, data: Any) -> List[List[Dict[str, float]]]:
26
+ inputs = data.pop("inputs", data)
27
+ parameters = data.pop("parameters", None)
28
+ if parameters is not None:
29
+ prediction = self.pipeline(inputs, **parameters)
30
+ else:
31
+ prediction = self.pipeline(inputs)
32
+ return prediction
requirements.txt ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ peft==0.9.0
2
+ transformers==4.38.2
runs/Mar21_02-37-23_r-nicolof88-mistral7b-spider-at-188n2lyh-a2f2a-vnlru/events.out.tfevents.1710988652.r-nicolof88-mistral7b-spider-at-188n2lyh-a2f2a-vnlru.98.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7612823918dcca134557e133967c50bf0947aba2a2cd56218766fee871b62c36
3
- size 86510
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a68f799653a345af0889cf7ca9e4f06849262070d3ead3f614d9ca0904a4d1b1
3
+ size 87075
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "additional_special_tokens": [],
31
+ "bos_token": "<s>",
32
+ "chat_template": "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ message['content'] + eos_token}}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}",
33
+ "clean_up_tokenization_spaces": false,
34
+ "eos_token": "</s>",
35
+ "legacy": true,
36
+ "model_max_length": 2048,
37
+ "pad_token": "</s>",
38
+ "sp_model_kwargs": {},
39
+ "spaces_between_special_tokens": false,
40
+ "tokenizer_class": "LlamaTokenizer",
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": false
43
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:471278576427115d6d17bfb25bab0702b85641a0b161bfd1e29e51fbba4b4223
3
+ size 4984
training_params.json ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "mistralai/Mistral-7B-Instruct-v0.2",
3
+ "project_name": "mistral7b-spider-ft",
4
+ "data_path": "mistral7b-spider-ft/autotrain-data",
5
+ "train_split": "train",
6
+ "valid_split": null,
7
+ "add_eos_token": true,
8
+ "block_size": 1024,
9
+ "model_max_length": 2048,
10
+ "padding": "right",
11
+ "trainer": "sft",
12
+ "use_flash_attention_2": false,
13
+ "log": "tensorboard",
14
+ "disable_gradient_checkpointing": false,
15
+ "logging_steps": -1,
16
+ "evaluation_strategy": "epoch",
17
+ "save_total_limit": 1,
18
+ "save_strategy": "epoch",
19
+ "auto_find_batch_size": false,
20
+ "mixed_precision": "fp16",
21
+ "lr": 3e-05,
22
+ "epochs": 4,
23
+ "batch_size": 2,
24
+ "warmup_ratio": 0.1,
25
+ "gradient_accumulation": 1,
26
+ "optimizer": "adamw_torch",
27
+ "scheduler": "linear",
28
+ "weight_decay": 0.0,
29
+ "max_grad_norm": 1.0,
30
+ "seed": 42,
31
+ "chat_template": "none",
32
+ "quantization": "int4",
33
+ "target_modules": "all-linear",
34
+ "merge_adapter": false,
35
+ "peft": true,
36
+ "lora_r": 64,
37
+ "lora_alpha": 16,
38
+ "lora_dropout": 0.05,
39
+ "model_ref": null,
40
+ "dpo_beta": 0.1,
41
+ "prompt_text_column": "autotrain_prompt",
42
+ "text_column": "autotrain_text",
43
+ "rejected_text_column": "autotrain_rejected_text",
44
+ "push_to_hub": true,
45
+ "repo_id": "nicolof88/mistral7b-spider-ft",
46
+ "username": "nicolof88"
47
+ }