Hugging Face's logo

wls04
/
gemma_web

Image-Text-to-Text
Transformers
TensorBoard
Safetensors
gemma3
llama-factory
full
Generated from Trainer
conversational
text-generation-inference
Model card Files
xet
 Metrics Community
wls04 commited on
Commit
20acb38
·
verified ·
1 Parent(s): 0de51be

Upload model

Browse files
Files changed (45) hide show
  1. .gitattributes +2 -0
  2. README.md +63 -0
  3. added_tokens.json +4 -0
  4. all_results.json +12 -0
  5. chat_template.jinja +47 -0
  6. checkpoint-67/added_tokens.json +4 -0
  7. checkpoint-67/chat_template.jinja +47 -0
  8. checkpoint-67/config.json +64 -0
  9. checkpoint-67/generation_config.json +13 -0
  10. checkpoint-67/model-00001-of-00005.safetensors +3 -0
  11. checkpoint-67/model-00002-of-00005.safetensors +3 -0
  12. checkpoint-67/model-00003-of-00005.safetensors +3 -0
  13. checkpoint-67/model-00004-of-00005.safetensors +3 -0
  14. checkpoint-67/model-00005-of-00005.safetensors +3 -0
  15. checkpoint-67/model.safetensors.index.json +0 -0
  16. checkpoint-67/preprocessor_config.json +29 -0
  17. checkpoint-67/processor_config.json +4 -0
  18. checkpoint-67/special_tokens_map.json +33 -0
  19. checkpoint-67/tokenizer.json +3 -0
  20. checkpoint-67/tokenizer.model +3 -0
  21. checkpoint-67/tokenizer_config.json +0 -0
  22. checkpoint-67/trainer_state.json +76 -0
  23. checkpoint-67/training_args.bin +3 -0
  24. config.json +64 -0
  25. eval_results.json +7 -0
  26. generation_config.json +13 -0
  27. model-00001-of-00005.safetensors +3 -0
  28. model-00002-of-00005.safetensors +3 -0
  29. model-00003-of-00005.safetensors +3 -0
  30. model-00004-of-00005.safetensors +3 -0
  31. model-00005-of-00005.safetensors +3 -0
  32. model.safetensors.index.json +0 -0
  33. preprocessor_config.json +29 -0
  34. processor_config.json +4 -0
  35. runs/Jun19_02-20-56_node4/events.out.tfevents.1750267456.node4.22007.0 +3 -0
  36. runs/Jun19_02-20-56_node4/events.out.tfevents.1750293237.node4.22007.1 +3 -0
  37. special_tokens_map.json +33 -0
  38. tokenizer.json +3 -0
  39. tokenizer.model +3 -0
  40. tokenizer_config.json +0 -0
  41. train_results.json +8 -0
  42. trainer_log.jsonl +7 -0
  43. trainer_state.json +85 -0
  44. training_args.bin +3 -0
  45. training_loss.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
+ checkpoint-67/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: other
4
+ base_model: google/gemma-3-12b-it
5
+ tags:
6
+ - llama-factory
7
+ - full
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: webrl_gemma
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # webrl_gemma
18
+
19
+ This model is a fine-tuned version of [google/gemma-3-12b-it](https://huggingface.co/google/gemma-3-12b-it) on the webrl_9k dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.1657
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 1e-05
41
+ - train_batch_size: 1
42
+ - eval_batch_size: 1
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - num_devices: 8
46
+ - gradient_accumulation_steps: 16
47
+ - total_train_batch_size: 128
48
+ - total_eval_batch_size: 8
49
+ - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
50
+ - lr_scheduler_type: cosine
51
+ - lr_scheduler_warmup_ratio: 0.1
52
+ - num_epochs: 1
53
+
54
+ ### Training results
55
+
56
+
57
+
58
+ ### Framework versions
59
+
60
+ - Transformers 4.52.4
61
+ - Pytorch 2.7.1+cu126
62
+ - Datasets 3.6.0
63
+ - Tokenizers 0.21.1
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<image_soft_token>": 262144,
3
+ "<|eot_id|>": 262145
4
+ }
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 0.16572383046150208,
4
+ "eval_runtime": 482.8437,
5
+ "eval_samples_per_second": 1.959,
6
+ "eval_steps_per_second": 0.246,
7
+ "total_flos": 265909239283712.0,
8
+ "train_loss": 0.48691165269310793,
9
+ "train_runtime": 25225.3606,
10
+ "train_samples_per_second": 0.338,
11
+ "train_steps_per_second": 0.003
12
+ }
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 -%}
checkpoint-67/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<image_soft_token>": 262144,
3
+ "<|eot_id|>": 262145
4
+ }
checkpoint-67/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 -%}
checkpoint-67/config.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "eoi_token_index": 256000,
7
+ "eos_token_id": [
8
+ 1,
9
+ 106
10
+ ],
11
+ "hidden_size": 3840,
12
+ "image_token_index": 262144,
13
+ "initializer_range": 0.02,
14
+ "mm_tokens_per_image": 256,
15
+ "model_type": "gemma3",
16
+ "text_config": {
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "cache_implementation": "hybrid",
21
+ "final_logit_softcapping": null,
22
+ "head_dim": 256,
23
+ "hidden_activation": "gelu_pytorch_tanh",
24
+ "hidden_size": 3840,
25
+ "initializer_range": 0.02,
26
+ "intermediate_size": 15360,
27
+ "max_position_embeddings": 131072,
28
+ "model_type": "gemma3_text",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 48,
31
+ "num_key_value_heads": 8,
32
+ "query_pre_attn_scalar": 256,
33
+ "rms_norm_eps": 1e-06,
34
+ "rope_local_base_freq": 10000.0,
35
+ "rope_scaling": {
36
+ "factor": 8.0,
37
+ "rope_type": "linear"
38
+ },
39
+ "rope_theta": 1000000.0,
40
+ "sliding_window": 1024,
41
+ "sliding_window_pattern": 6,
42
+ "torch_dtype": "float32",
43
+ "use_cache": false,
44
+ "vocab_size": 262208
45
+ },
46
+ "torch_dtype": "bfloat16",
47
+ "transformers_version": "4.52.4",
48
+ "use_cache": false,
49
+ "vision_config": {
50
+ "attention_dropout": 0.0,
51
+ "hidden_act": "gelu_pytorch_tanh",
52
+ "hidden_size": 1152,
53
+ "image_size": 896,
54
+ "intermediate_size": 4304,
55
+ "layer_norm_eps": 1e-06,
56
+ "model_type": "siglip_vision_model",
57
+ "num_attention_heads": 16,
58
+ "num_channels": 3,
59
+ "num_hidden_layers": 27,
60
+ "patch_size": 14,
61
+ "torch_dtype": "float32",
62
+ "vision_use_head": false
63
+ }
64
+ }
checkpoint-67/generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 106
8
+ ],
9
+ "pad_token_id": 0,
10
+ "top_k": 64,
11
+ "top_p": 0.95,
12
+ "transformers_version": "4.52.4"
13
+ }
checkpoint-67/model-00001-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7501dd55ea3cfeafc0b71c6aa051e7a15a16ba816eff1e91e2c8fbc821cf5bf9
3
+ size 4979902192
checkpoint-67/model-00002-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05d8f625fc8aee94669f7c174bfc084b907f19c53c2136b8b06c23a42976f847
3
+ size 4931296592
checkpoint-67/model-00003-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b4e087a26a73b59259cf0a78eff0639cfc307fdf1eb37c5d917f1494d558c55
3
+ size 4931296656
checkpoint-67/model-00004-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1b43670c43881f442cf692a081217b8578534d9a1d892a9d9e61bc0e8742803
3
+ size 4931296656
checkpoint-67/model-00005-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c9f8dc079195c8019417a3bb6735005f31ff3c492c11aea47bb18d956a012d4
3
+ size 4601000928
checkpoint-67/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-67/preprocessor_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_rgb": null,
3
+ "do_normalize": true,
4
+ "do_pan_and_scan": null,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "image_mean": [
8
+ 0.5,
9
+ 0.5,
10
+ 0.5
11
+ ],
12
+ "image_processor_type": "Gemma3ImageProcessor",
13
+ "image_seq_length": 256,
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "pan_and_scan_max_num_crops": null,
20
+ "pan_and_scan_min_crop_size": null,
21
+ "pan_and_scan_min_ratio_to_activate": null,
22
+ "processor_class": "Gemma3Processor",
23
+ "resample": 2,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "height": 896,
27
+ "width": 896
28
+ }
29
+ }
checkpoint-67/processor_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "image_seq_length": 256,
3
+ "processor_class": "Gemma3Processor"
4
+ }
checkpoint-67/special_tokens_map.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "boi_token": "<start_of_image>",
3
+ "bos_token": {
4
+ "content": "<bos>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "eoi_token": "<end_of_image>",
11
+ "eos_token": {
12
+ "content": "<|eot_id|>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "image_token": "<image_soft_token>",
19
+ "pad_token": {
20
+ "content": "<pad>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false
25
+ },
26
+ "unk_token": {
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ }
33
+ }
checkpoint-67/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee06f53dcc483d8da3793a84ab3137de857b52f9226315d2aa3172a0c5769975
3
+ size 33384755
checkpoint-67/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
3
+ size 4689074
checkpoint-67/tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-67/trainer_state.json ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0,
6
+ "eval_steps": 1000,
7
+ "global_step": 67,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.15023474178403756,
14
+ "grad_norm": 31.56661033630371,
15
+ "learning_rate": 9.972609476841368e-06,
16
+ "loss": 1.7659,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.3004694835680751,
21
+ "grad_norm": 35.23997497558594,
22
+ "learning_rate": 9.045084971874738e-06,
23
+ "loss": 0.5183,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.4507042253521127,
28
+ "grad_norm": 5.03916072845459,
29
+ "learning_rate": 7.033683215379002e-06,
30
+ "loss": 0.278,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.6009389671361502,
35
+ "grad_norm": 3.3426177501678467,
36
+ "learning_rate": 4.477357683661734e-06,
37
+ "loss": 0.2194,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.7511737089201878,
42
+ "grad_norm": 3.752589702606201,
43
+ "learning_rate": 2.061073738537635e-06,
44
+ "loss": 0.1969,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.9014084507042254,
49
+ "grad_norm": 5.254862308502197,
50
+ "learning_rate": 4.322727117869951e-07,
51
+ "loss": 0.1813,
52
+ "step": 60
53
+ }
54
+ ],
55
+ "logging_steps": 10,
56
+ "max_steps": 67,
57
+ "num_input_tokens_seen": 0,
58
+ "num_train_epochs": 1,
59
+ "save_steps": 100,
60
+ "stateful_callbacks": {
61
+ "TrainerControl": {
62
+ "args": {
63
+ "should_epoch_stop": false,
64
+ "should_evaluate": false,
65
+ "should_log": false,
66
+ "should_save": true,
67
+ "should_training_stop": true
68
+ },
69
+ "attributes": {}
70
+ }
71
+ },
72
+ "total_flos": 265909239283712.0,
73
+ "train_batch_size": 1,
74
+ "trial_name": null,
75
+ "trial_params": null
76
+ }
checkpoint-67/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f36885d377f4ba39710b01513a039ea0c522a4fecc325704e143755fd3e9d426
3
+ size 8145
config.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "eoi_token_index": 256000,
7
+ "eos_token_id": [
8
+ 1,
9
+ 106
10
+ ],
11
+ "hidden_size": 3840,
12
+ "image_token_index": 262144,
13
+ "initializer_range": 0.02,
14
+ "mm_tokens_per_image": 256,
15
+ "model_type": "gemma3",
16
+ "text_config": {
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "cache_implementation": "hybrid",
21
+ "final_logit_softcapping": null,
22
+ "head_dim": 256,
23
+ "hidden_activation": "gelu_pytorch_tanh",
24
+ "hidden_size": 3840,
25
+ "initializer_range": 0.02,
26
+ "intermediate_size": 15360,
27
+ "max_position_embeddings": 131072,
28
+ "model_type": "gemma3_text",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 48,
31
+ "num_key_value_heads": 8,
32
+ "query_pre_attn_scalar": 256,
33
+ "rms_norm_eps": 1e-06,
34
+ "rope_local_base_freq": 10000.0,
35
+ "rope_scaling": {
36
+ "factor": 8.0,
37
+ "rope_type": "linear"
38
+ },
39
+ "rope_theta": 1000000.0,
40
+ "sliding_window": 1024,
41
+ "sliding_window_pattern": 6,
42
+ "torch_dtype": "float32",
43
+ "use_cache": false,
44
+ "vocab_size": 262208
45
+ },
46
+ "torch_dtype": "bfloat16",
47
+ "transformers_version": "4.52.4",
48
+ "use_cache": false,
49
+ "vision_config": {
50
+ "attention_dropout": 0.0,
51
+ "hidden_act": "gelu_pytorch_tanh",
52
+ "hidden_size": 1152,
53
+ "image_size": 896,
54
+ "intermediate_size": 4304,
55
+ "layer_norm_eps": 1e-06,
56
+ "model_type": "siglip_vision_model",
57
+ "num_attention_heads": 16,
58
+ "num_channels": 3,
59
+ "num_hidden_layers": 27,
60
+ "patch_size": 14,
61
+ "torch_dtype": "float32",
62
+ "vision_use_head": false
63
+ }
64
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 0.16572383046150208,
4
+ "eval_runtime": 482.8437,
5
+ "eval_samples_per_second": 1.959,
6
+ "eval_steps_per_second": 0.246
7
+ }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 106
8
+ ],
9
+ "pad_token_id": 0,
10
+ "top_k": 64,
11
+ "top_p": 0.95,
12
+ "transformers_version": "4.52.4"
13
+ }
model-00001-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7501dd55ea3cfeafc0b71c6aa051e7a15a16ba816eff1e91e2c8fbc821cf5bf9
3
+ size 4979902192
model-00002-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05d8f625fc8aee94669f7c174bfc084b907f19c53c2136b8b06c23a42976f847
3
+ size 4931296592
model-00003-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b4e087a26a73b59259cf0a78eff0639cfc307fdf1eb37c5d917f1494d558c55
3
+ size 4931296656
model-00004-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1b43670c43881f442cf692a081217b8578534d9a1d892a9d9e61bc0e8742803
3
+ size 4931296656
model-00005-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c9f8dc079195c8019417a3bb6735005f31ff3c492c11aea47bb18d956a012d4
3
+ size 4601000928
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_rgb": null,
3
+ "do_normalize": true,
4
+ "do_pan_and_scan": null,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "image_mean": [
8
+ 0.5,
9
+ 0.5,
10
+ 0.5
11
+ ],
12
+ "image_processor_type": "Gemma3ImageProcessor",
13
+ "image_seq_length": 256,
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "pan_and_scan_max_num_crops": null,
20
+ "pan_and_scan_min_crop_size": null,
21
+ "pan_and_scan_min_ratio_to_activate": null,
22
+ "processor_class": "Gemma3Processor",
23
+ "resample": 2,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "height": 896,
27
+ "width": 896
28
+ }
29
+ }
processor_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "image_seq_length": 256,
3
+ "processor_class": "Gemma3Processor"
4
+ }
runs/Jun19_02-20-56_node4/events.out.tfevents.1750267456.node4.22007.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0c3e4c22dd6f13c8b11072fe330d17089aabd30b275917c08854b4f2d5b9143
3
+ size 8024
runs/Jun19_02-20-56_node4/events.out.tfevents.1750293237.node4.22007.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:586a523facbf38f27da6368054e6637b4c32a09f7ab02d32fb027ff5d8da3091
3
+ size 354
special_tokens_map.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "boi_token": "<start_of_image>",
3
+ "bos_token": {
4
+ "content": "<bos>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "eoi_token": "<end_of_image>",
11
+ "eos_token": {
12
+ "content": "<|eot_id|>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "image_token": "<image_soft_token>",
19
+ "pad_token": {
20
+ "content": "<pad>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false
25
+ },
26
+ "unk_token": {
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ }
33
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee06f53dcc483d8da3793a84ab3137de857b52f9226315d2aa3172a0c5769975
3
+ size 33384755
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
3
+ size 4689074
tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff
 
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "total_flos": 265909239283712.0,
4
+ "train_loss": 0.48691165269310793,
5
+ "train_runtime": 25225.3606,
6
+ "train_samples_per_second": 0.338,
7
+ "train_steps_per_second": 0.003
8
+ }
trainer_log.jsonl ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {"current_steps": 10, "total_steps": 67, "loss": 1.7659, "lr": 9.972609476841368e-06, "epoch": 0.15023474178403756, "percentage": 14.93, "elapsed_time": "1:02:40", "remaining_time": "5:57:17"}
2
+ {"current_steps": 20, "total_steps": 67, "loss": 0.5183, "lr": 9.045084971874738e-06, "epoch": 0.3004694835680751, "percentage": 29.85, "elapsed_time": "2:05:42", "remaining_time": "4:55:25"}
3
+ {"current_steps": 30, "total_steps": 67, "loss": 0.278, "lr": 7.033683215379002e-06, "epoch": 0.4507042253521127, "percentage": 44.78, "elapsed_time": "3:07:59", "remaining_time": "3:51:50"}
4
+ {"current_steps": 40, "total_steps": 67, "loss": 0.2194, "lr": 4.477357683661734e-06, "epoch": 0.6009389671361502, "percentage": 59.7, "elapsed_time": "4:10:01", "remaining_time": "2:48:45"}
5
+ {"current_steps": 50, "total_steps": 67, "loss": 0.1969, "lr": 2.061073738537635e-06, "epoch": 0.7511737089201878, "percentage": 74.63, "elapsed_time": "5:13:51", "remaining_time": "1:46:42"}
6
+ {"current_steps": 60, "total_steps": 67, "loss": 0.1813, "lr": 4.322727117869951e-07, "epoch": 0.9014084507042254, "percentage": 89.55, "elapsed_time": "6:18:20", "remaining_time": "0:44:08"}
7
+ {"current_steps": 67, "total_steps": 67, "epoch": 1.0, "percentage": 100.0, "elapsed_time": "7:00:25", "remaining_time": "0:00:00"}
trainer_state.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0,
6
+ "eval_steps": 1000,
7
+ "global_step": 67,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.15023474178403756,
14
+ "grad_norm": 31.56661033630371,
15
+ "learning_rate": 9.972609476841368e-06,
16
+ "loss": 1.7659,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.3004694835680751,
21
+ "grad_norm": 35.23997497558594,
22
+ "learning_rate": 9.045084971874738e-06,
23
+ "loss": 0.5183,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.4507042253521127,
28
+ "grad_norm": 5.03916072845459,
29
+ "learning_rate": 7.033683215379002e-06,
30
+ "loss": 0.278,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.6009389671361502,
35
+ "grad_norm": 3.3426177501678467,
36
+ "learning_rate": 4.477357683661734e-06,
37
+ "loss": 0.2194,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.7511737089201878,
42
+ "grad_norm": 3.752589702606201,
43
+ "learning_rate": 2.061073738537635e-06,
44
+ "loss": 0.1969,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.9014084507042254,
49
+ "grad_norm": 5.254862308502197,
50
+ "learning_rate": 4.322727117869951e-07,
51
+ "loss": 0.1813,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "step": 67,
57
+ "total_flos": 265909239283712.0,
58
+ "train_loss": 0.48691165269310793,
59
+ "train_runtime": 25225.3606,
60
+ "train_samples_per_second": 0.338,
61
+ "train_steps_per_second": 0.003
62
+ }
63
+ ],
64
+ "logging_steps": 10,
65
+ "max_steps": 67,
66
+ "num_input_tokens_seen": 0,
67
+ "num_train_epochs": 1,
68
+ "save_steps": 100,
69
+ "stateful_callbacks": {
70
+ "TrainerControl": {
71
+ "args": {
72
+ "should_epoch_stop": false,
73
+ "should_evaluate": false,
74
+ "should_log": false,
75
+ "should_save": true,
76
+ "should_training_stop": true
77
+ },
78
+ "attributes": {}
79
+ }
80
+ },
81
+ "total_flos": 265909239283712.0,
82
+ "train_batch_size": 1,
83
+ "trial_name": null,
84
+ "trial_params": null
85
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f36885d377f4ba39710b01513a039ea0c522a4fecc325704e143755fd3e9d426
3
+ size 8145
training_loss.png ADDED