Hugging Face's logo

Oltking
/
adaption_agronomy_problem_solving

PEFT
Safetensors
gemma3_text
lora
adapter
adaption
Model card Files
xet
 Community
Oltking commited on
Commit
f82e40d
·
verified ·
1 Parent(s): 195fe50

Add 11 files

Browse files
Files changed (12) hide show
  1. .gitattributes +2 -0
  2. README.md +99 -0
  3. adapter_config.json +39 -0
  4. adapter_model.safetensors +3 -0
  5. chat_template.jinja +47 -0
  6. config.json +81 -0
  7. special_tokens_map.json +10 -0
  8. tokenizer.json +3 -0
  9. tokenizer_config.json +26 -0
  10. trainer_state.json +305 -0
  11. training-metrics.png +3 -0
  12. win-rates.png +0 -0
.gitattributes CHANGED
@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ training-metrics.png filter=lfs diff=lfs merge=lfs -text
37
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: google/gemma-3-4b-it
3
+ library_name: peft
4
+ license: other
5
+ tags:
6
+ - lora
7
+ - peft
8
+ - adapter
9
+ - adaption
10
+ ---
11
+
12
+ # adaption_agronomy_problem_solving
13
+
14
+ ## Model Training
15
+
16
+ A LORA adapter for `google/gemma-3-4b-it`. This model was trained with SFT using [Adaption](https://adaptionlabs.ai)'s AutoScientist on the agronomy_problem_solving dataset.
17
+
18
+
19
+ ![Training metrics](training-metrics.png)
20
+
21
+ ### AutoScientist Config
22
+
23
+ ```json
24
+ {
25
+ "job_id": "18fe9ddd-e21b-43b3-8739-c07511cd5337",
26
+ "training_experiment_id": "0352f9e2-8e81-4302-8d91-5d0eee1eb330",
27
+ "original_model_name": "google/gemma-3-4b-it",
28
+ "trained_model_name": "adaption_agronomy_problem_solving",
29
+ "training_method": "sft",
30
+ "training_type": "lora",
31
+ "data_format": "chat",
32
+ "hyperparams": {
33
+ "lora": "true",
34
+ "lora_r": 16,
35
+ "n_evals": 5,
36
+ "n_epochs": 1,
37
+ "batch_size": "max",
38
+ "lora_alpha": 32,
39
+ "lora_dropout": 0,
40
+ "min_lr_ratio": 0.1,
41
+ "warmup_ratio": 0.1,
42
+ "weight_decay": 0,
43
+ "learning_rate": 0.00001,
44
+ "max_grad_norm": 2,
45
+ "base_model_size": "4B",
46
+ "train_on_inputs": "false",
47
+ "training_method": "sft",
48
+ "lr_scheduler_type": "cosine",
49
+ "scheduler_num_cycles": 0.5,
50
+ "lora_trainable_modules": "all-linear"
51
+ }
52
+ }
53
+ ```
54
+
55
+ ## Training Data
56
+
57
+ The model was trained on 27,622 rows of adapted data with the following domain distribution: agriculture (97%), math (3%).
58
+
59
+ ## Model Evaluation
60
+
61
+ The model was evaluated on an in-distribution held-out test set as well as a broader domain-specific test set to measure generalization.
62
+
63
+
64
+ ![Win rates](win-rates.png)
65
+
66
+
67
+ ## How to use
68
+
69
+ ```bash
70
+ pip install torch transformers peft
71
+ ```
72
+
73
+ ```python
74
+ import torch
75
+ from transformers import AutoModelForCausalLM, AutoTokenizer
76
+ from peft import PeftModel
77
+
78
+ BASE = "google/gemma-3-4b-it"
79
+ ADAPTER = "<this-repo-id>"
80
+
81
+ device = "cuda" if torch.cuda.is_available() else "cpu"
82
+ dtype = torch.float32 if device == "cpu" else torch.bfloat16
83
+
84
+ base = AutoModelForCausalLM.from_pretrained(BASE, dtype=dtype).to(device)
85
+ model = PeftModel.from_pretrained(base, ADAPTER)
86
+ # Optional: merge the LoRA weights into the base for faster inference
87
+ model = model.merge_and_unload()
88
+ model.eval()
89
+
90
+ tokenizer = AutoTokenizer.from_pretrained(BASE)
91
+ messages = [{"role": "user", "content": "Hello!"}]
92
+ text = tokenizer.apply_chat_template(
93
+ messages, tokenize=False, add_generation_prompt=True)
94
+ inputs = tokenizer(text, return_tensors="pt").to(device)
95
+
96
+ with torch.inference_mode():
97
+ out = model.generate(**inputs, max_new_tokens=512)
98
+ print(tokenizer.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
99
+ ```
adapter_config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "togethercomputer/gemma-3-4b-it",
5
+ "bias": "none",
6
+ "corda_config": null,
7
+ "eva_config": null,
8
+ "exclude_modules": [],
9
+ "fan_in_fan_out": false,
10
+ "inference_mode": true,
11
+ "init_lora_weights": true,
12
+ "layer_replication": null,
13
+ "layers_pattern": null,
14
+ "layers_to_transform": null,
15
+ "loftq_config": {},
16
+ "lora_alpha": 32,
17
+ "lora_bias": false,
18
+ "lora_dropout": 0.0,
19
+ "megatron_config": null,
20
+ "megatron_core": "megatron.core",
21
+ "modules_to_save": null,
22
+ "peft_type": "LORA",
23
+ "r": 16,
24
+ "rank_pattern": {},
25
+ "revision": null,
26
+ "target_modules": [
27
+ "k_proj",
28
+ "q_proj",
29
+ "gate_proj",
30
+ "down_proj",
31
+ "up_proj",
32
+ "v_proj",
33
+ "o_proj"
34
+ ],
35
+ "task_type": "CAUSAL_LM",
36
+ "trainable_token_indices": null,
37
+ "use_dora": false,
38
+ "use_rslora": false
39
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1274c7fa8c7de520cb81103eb90b9b2f4de2cba01cb02e23846a28284d499b2a
3
+ size 119273568
chat_template.jinja ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {{ bos_token }}
2
+ {%- if messages[0]['role'] == 'system' -%}
3
+ {%- if messages[0]['content'] is string -%}
4
+ {%- set first_user_prefix = messages[0]['content'] + '
5
+
6
+ ' -%}
7
+ {%- else -%}
8
+ {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
9
+
10
+ ' -%}
11
+ {%- endif -%}
12
+ {%- set loop_messages = messages[1:] -%}
13
+ {%- else -%}
14
+ {%- set first_user_prefix = "" -%}
15
+ {%- set loop_messages = messages -%}
16
+ {%- endif -%}
17
+ {%- for message in loop_messages -%}
18
+ {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
19
+ {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
20
+ {%- endif -%}
21
+ {%- if (message['role'] == 'assistant') -%}
22
+ {%- set role = "model" -%}
23
+ {%- else -%}
24
+ {%- set role = message['role'] -%}
25
+ {%- endif -%}
26
+ {{ '<start_of_turn>' + role + '
27
+ ' + (first_user_prefix if loop.first else "") }}
28
+ {%- if message['content'] is string -%}
29
+ {{ message['content'] | trim }}
30
+ {%- elif message['content'] is iterable -%}
31
+ {%- for item in message['content'] -%}
32
+ {%- if item['type'] == 'image' -%}
33
+ {{ '<start_of_image>' }}
34
+ {%- elif item['type'] == 'text' -%}
35
+ {{ item['text'] | trim }}
36
+ {%- endif -%}
37
+ {%- endfor -%}
38
+ {%- else -%}
39
+ {{ raise_exception("Invalid content type") }}
40
+ {%- endif -%}
41
+ {{ '<end_of_turn>
42
+ ' }}
43
+ {%- endfor -%}
44
+ {%- if add_generation_prompt -%}
45
+ {{'<start_of_turn>model
46
+ '}}
47
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_sliding_window_pattern": 6,
3
+ "architectures": [
4
+ "Gemma3ForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "attn_logit_softcapping": null,
9
+ "bos_token_id": 2,
10
+ "dtype": "bfloat16",
11
+ "eos_token_id": 1,
12
+ "final_logit_softcapping": null,
13
+ "head_dim": 256,
14
+ "hidden_activation": "gelu_pytorch_tanh",
15
+ "hidden_size": 2560,
16
+ "initializer_range": 0.02,
17
+ "intermediate_size": 10240,
18
+ "layer_types": [
19
+ "sliding_attention",
20
+ "sliding_attention",
21
+ "sliding_attention",
22
+ "sliding_attention",
23
+ "sliding_attention",
24
+ "full_attention",
25
+ "sliding_attention",
26
+ "sliding_attention",
27
+ "sliding_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "full_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "full_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "sliding_attention",
42
+ "full_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "sliding_attention",
48
+ "full_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "sliding_attention",
52
+ "sliding_attention"
53
+ ],
54
+ "max_position_embeddings": 131072,
55
+ "mm_tokens_per_image": 256,
56
+ "model_type": "gemma3_text",
57
+ "num_attention_heads": 8,
58
+ "num_hidden_layers": 34,
59
+ "num_key_value_heads": 4,
60
+ "pad_token_id": 0,
61
+ "query_pre_attn_scalar": 256,
62
+ "rms_norm_eps": 1e-06,
63
+ "rope_parameters": {
64
+ "full_attention": {
65
+ "factor": 8.0,
66
+ "rope_theta": 1000000.0,
67
+ "rope_type": "linear"
68
+ },
69
+ "sliding_attention": {
70
+ "rope_theta": 10000.0,
71
+ "rope_type": "default"
72
+ }
73
+ },
74
+ "sliding_window": 1024,
75
+ "tie_word_embeddings": true,
76
+ "transformers_version": "5.10.1",
77
+ "use_bidirectional_attention": false,
78
+ "use_cache": false,
79
+ "vocab_size": 262208,
80
+ "torch_dtype": "bfloat16"
81
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "boi_token": "<start_of_image>",
3
+ "bos_token": "<bos>",
4
+ "eoi_token": "<end_of_image>",
5
+ "eos_token": "<eos>",
6
+ "image_token": "<image_soft_token>",
7
+ "mask_token": "<mask>",
8
+ "pad_token": "<pad>",
9
+ "unk_token": "<unk>"
10
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
3
+ size 33384567
tokenizer_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "boi_token": "<start_of_image>",
4
+ "bos_token": "<bos>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eoi_token": "<end_of_image>",
7
+ "eos_token": "<eos>",
8
+ "image_token": "<image_soft_token>",
9
+ "is_local": false,
10
+ "local_files_only": true,
11
+ "mask_token": "<mask>",
12
+ "model_max_length": 131072,
13
+ "model_specific_special_tokens": {
14
+ "boi_token": "<start_of_image>",
15
+ "eoi_token": "<end_of_image>",
16
+ "image_token": "<image_soft_token>"
17
+ },
18
+ "pad_token": "<pad>",
19
+ "padding_side": "right",
20
+ "processor_class": "Gemma3Processor",
21
+ "sp_model_kwargs": null,
22
+ "spaces_between_special_tokens": false,
23
+ "tokenizer_class": "GemmaTokenizer",
24
+ "unk_token": "<unk>",
25
+ "use_default_system_prompt": false
26
+ }
trainer_state.json ADDED
@@ -0,0 +1,305 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0,
6
+ "eval_steps": 6,
7
+ "global_step": 33,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.030303030303030304,
14
+ "grad_norm": 1.4287192821502686,
15
+ "learning_rate": 0.0,
16
+ "loss": 2.0361328125,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.06060606060606061,
21
+ "grad_norm": 1.437935471534729,
22
+ "learning_rate": 2.5e-06,
23
+ "loss": 2.1640625,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.09090909090909091,
28
+ "grad_norm": 1.3686531782150269,
29
+ "learning_rate": 5e-06,
30
+ "loss": 2.01171875,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.12121212121212122,
35
+ "grad_norm": 1.286812424659729,
36
+ "learning_rate": 7.500000000000001e-06,
37
+ "loss": 1.9638671875,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.15151515151515152,
42
+ "grad_norm": 1.0939735174179077,
43
+ "learning_rate": 1e-05,
44
+ "loss": 1.8427734375,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.18181818181818182,
49
+ "grad_norm": 1.7438583374023438,
50
+ "learning_rate": 9.97362080719462e-06,
51
+ "loss": 2.0869140625,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.21212121212121213,
56
+ "grad_norm": 0.9894726872444153,
57
+ "learning_rate": 9.89479250069539e-06,
58
+ "loss": 1.76953125,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.24242424242424243,
63
+ "grad_norm": 0.8804801106452942,
64
+ "learning_rate": 9.764439270322612e-06,
65
+ "loss": 1.7158203125,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.2727272727272727,
70
+ "grad_norm": 0.6704576015472412,
71
+ "learning_rate": 9.584089388519307e-06,
72
+ "loss": 1.7607421875,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.2727272727272727,
77
+ "eval_loss": 2.0078125,
78
+ "eval_runtime": 2.5535,
79
+ "eval_samples_per_second": 0.783,
80
+ "eval_steps_per_second": 0.392,
81
+ "step": 9
82
+ },
83
+ {
84
+ "epoch": 0.30303030303030304,
85
+ "grad_norm": 1.9209635257720947,
86
+ "learning_rate": 9.355857292754152e-06,
87
+ "loss": 2.0673828125,
88
+ "step": 10
89
+ },
90
+ {
91
+ "epoch": 0.3333333333333333,
92
+ "grad_norm": 0.8329216837882996,
93
+ "learning_rate": 9.082418795675397e-06,
94
+ "loss": 1.677734375,
95
+ "step": 11
96
+ },
97
+ {
98
+ "epoch": 0.36363636363636365,
99
+ "grad_norm": 0.9212274551391602,
100
+ "learning_rate": 8.76697971365409e-06,
101
+ "loss": 1.9130859375,
102
+ "step": 12
103
+ },
104
+ {
105
+ "epoch": 0.3939393939393939,
106
+ "grad_norm": 1.0851582288742065,
107
+ "learning_rate": 8.413238281518225e-06,
108
+ "loss": 1.873046875,
109
+ "step": 13
110
+ },
111
+ {
112
+ "epoch": 0.42424242424242425,
113
+ "grad_norm": 0.9320775270462036,
114
+ "learning_rate": 8.025341794130722e-06,
115
+ "loss": 1.7841796875,
116
+ "step": 14
117
+ },
118
+ {
119
+ "epoch": 0.45454545454545453,
120
+ "grad_norm": 0.6170089244842529,
121
+ "learning_rate": 7.607837983149057e-06,
122
+ "loss": 1.5927734375,
123
+ "step": 15
124
+ },
125
+ {
126
+ "epoch": 0.45454545454545453,
127
+ "eval_loss": 1.8125,
128
+ "eval_runtime": 2.5501,
129
+ "eval_samples_per_second": 0.784,
130
+ "eval_steps_per_second": 0.392,
131
+ "step": 15
132
+ },
133
+ {
134
+ "epoch": 0.48484848484848486,
135
+ "grad_norm": 0.7249671816825867,
136
+ "learning_rate": 7.165621699029615e-06,
137
+ "loss": 1.7607421875,
138
+ "step": 16
139
+ },
140
+ {
141
+ "epoch": 0.5151515151515151,
142
+ "grad_norm": 0.6273460984230042,
143
+ "learning_rate": 6.703877523381495e-06,
144
+ "loss": 1.666015625,
145
+ "step": 17
146
+ },
147
+ {
148
+ "epoch": 0.5454545454545454,
149
+ "grad_norm": 0.6081604957580566,
150
+ "learning_rate": 6.228018984487443e-06,
151
+ "loss": 1.486328125,
152
+ "step": 18
153
+ },
154
+ {
155
+ "epoch": 0.5757575757575758,
156
+ "grad_norm": 0.8183671832084656,
157
+ "learning_rate": 5.74362508863438e-06,
158
+ "loss": 1.7958984375,
159
+ "step": 19
160
+ },
161
+ {
162
+ "epoch": 0.6060606060606061,
163
+ "grad_norm": 0.8645656704902649,
164
+ "learning_rate": 5.256374911365621e-06,
165
+ "loss": 1.6435546875,
166
+ "step": 20
167
+ },
168
+ {
169
+ "epoch": 0.6363636363636364,
170
+ "grad_norm": 0.7197942137718201,
171
+ "learning_rate": 4.771981015512559e-06,
172
+ "loss": 1.6796875,
173
+ "step": 21
174
+ },
175
+ {
176
+ "epoch": 0.6363636363636364,
177
+ "eval_loss": 1.703125,
178
+ "eval_runtime": 2.5572,
179
+ "eval_samples_per_second": 0.782,
180
+ "eval_steps_per_second": 0.391,
181
+ "step": 21
182
+ },
183
+ {
184
+ "epoch": 0.6666666666666666,
185
+ "grad_norm": 0.6852272152900696,
186
+ "learning_rate": 4.296122476618507e-06,
187
+ "loss": 1.71484375,
188
+ "step": 22
189
+ },
190
+ {
191
+ "epoch": 0.696969696969697,
192
+ "grad_norm": 0.5617450475692749,
193
+ "learning_rate": 3.834378300970385e-06,
194
+ "loss": 1.509765625,
195
+ "step": 23
196
+ },
197
+ {
198
+ "epoch": 0.7272727272727273,
199
+ "grad_norm": 0.7030121684074402,
200
+ "learning_rate": 3.392162016850945e-06,
201
+ "loss": 1.587890625,
202
+ "step": 24
203
+ },
204
+ {
205
+ "epoch": 0.7575757575757576,
206
+ "grad_norm": 0.5215224623680115,
207
+ "learning_rate": 2.9746582058692803e-06,
208
+ "loss": 1.5263671875,
209
+ "step": 25
210
+ },
211
+ {
212
+ "epoch": 0.7878787878787878,
213
+ "grad_norm": 0.5175567269325256,
214
+ "learning_rate": 2.586761718481776e-06,
215
+ "loss": 1.5361328125,
216
+ "step": 26
217
+ },
218
+ {
219
+ "epoch": 0.8181818181818182,
220
+ "grad_norm": 0.505078136920929,
221
+ "learning_rate": 2.2330202863459123e-06,
222
+ "loss": 1.5234375,
223
+ "step": 27
224
+ },
225
+ {
226
+ "epoch": 0.8181818181818182,
227
+ "eval_loss": 1.6484375,
228
+ "eval_runtime": 2.5566,
229
+ "eval_samples_per_second": 0.782,
230
+ "eval_steps_per_second": 0.391,
231
+ "step": 27
232
+ },
233
+ {
234
+ "epoch": 0.8484848484848485,
235
+ "grad_norm": 0.4583703279495239,
236
+ "learning_rate": 1.9175812043246034e-06,
237
+ "loss": 1.517578125,
238
+ "step": 28
239
+ },
240
+ {
241
+ "epoch": 0.8787878787878788,
242
+ "grad_norm": 0.47751930356025696,
243
+ "learning_rate": 1.6441427072458493e-06,
244
+ "loss": 1.5185546875,
245
+ "step": 29
246
+ },
247
+ {
248
+ "epoch": 0.9090909090909091,
249
+ "grad_norm": 0.5964409708976746,
250
+ "learning_rate": 1.4159106114806943e-06,
251
+ "loss": 1.4677734375,
252
+ "step": 30
253
+ },
254
+ {
255
+ "epoch": 0.9393939393939394,
256
+ "grad_norm": 0.4563347399234772,
257
+ "learning_rate": 1.2355607296773896e-06,
258
+ "loss": 1.54833984375,
259
+ "step": 31
260
+ },
261
+ {
262
+ "epoch": 0.9696969696969697,
263
+ "grad_norm": 0.48352736234664917,
264
+ "learning_rate": 1.1052074993046102e-06,
265
+ "loss": 1.44287109375,
266
+ "step": 32
267
+ },
268
+ {
269
+ "epoch": 1.0,
270
+ "grad_norm": 0.5558485984802246,
271
+ "learning_rate": 1.026379192805382e-06,
272
+ "loss": 1.48291015625,
273
+ "step": 33
274
+ },
275
+ {
276
+ "epoch": 1.0,
277
+ "eval_loss": 1.62109375,
278
+ "eval_runtime": 2.5478,
279
+ "eval_samples_per_second": 0.785,
280
+ "eval_steps_per_second": 0.393,
281
+ "step": 33
282
+ }
283
+ ],
284
+ "logging_steps": 1.0,
285
+ "max_steps": 33,
286
+ "num_input_tokens_seen": 0,
287
+ "num_train_epochs": 1,
288
+ "save_steps": 0,
289
+ "stateful_callbacks": {
290
+ "TrainerControl": {
291
+ "args": {
292
+ "should_epoch_stop": false,
293
+ "should_evaluate": false,
294
+ "should_log": false,
295
+ "should_save": true,
296
+ "should_training_stop": true
297
+ },
298
+ "attributes": {}
299
+ }
300
+ },
301
+ "total_flos": 6.715828683292017e+17,
302
+ "train_batch_size": 1,
303
+ "trial_name": null,
304
+ "trial_params": null
305
+ }
training-metrics.png ADDED

Git LFS Details

  • SHA256: d70e45cdc48594503a447132efdc0411167e0295eabf47dd1a2b9b69d8fec512
  • Pointer size: 131 Bytes
  • Size of remote file: 189 kB
win-rates.png ADDED