Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-g2_3_plan_oldcode_newcode

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
168e068
·
verified ·
1 Parent(s): 33b6dc9

Upload folder using huggingface_hub

Browse files
Files changed (24) hide show
  1. .gitattributes +1 -0
  2. 200M/chat_template.jinja +1 -0
  3. 200M/config.json +170 -0
  4. 200M/generation_config.json +17 -0
  5. 200M/latest +1 -0
  6. 200M/model.safetensors +3 -0
  7. 200M/processor_config.json +62 -0
  8. 200M/rng_state_0.pth +3 -0
  9. 200M/rng_state_1.pth +3 -0
  10. 200M/rng_state_10.pth +3 -0
  11. 200M/rng_state_11.pth +3 -0
  12. 200M/rng_state_2.pth +3 -0
  13. 200M/rng_state_3.pth +3 -0
  14. 200M/rng_state_4.pth +3 -0
  15. 200M/rng_state_5.pth +3 -0
  16. 200M/rng_state_6.pth +3 -0
  17. 200M/rng_state_7.pth +3 -0
  18. 200M/rng_state_8.pth +3 -0
  19. 200M/rng_state_9.pth +3 -0
  20. 200M/scheduler.pt +3 -0
  21. 200M/tokenizer.json +3 -0
  22. 200M/tokenizer_config.json +56 -0
  23. 200M/trainer_state.json +1455 -0
  24. 200M/training_args.bin +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ 200M/tokenizer.json filter=lfs diff=lfs merge=lfs -text
200M/chat_template.jinja ADDED
@@ -0,0 +1 @@
 
 
1
+ {% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ content }}{% elif message['role'] == 'assistant' %}{{ content }}{% endif %}{% endfor %}
200M/config.json ADDED
@@ -0,0 +1,170 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma4UnifiedForConditionalGeneration"
4
+ ],
5
+ "audio_config": {
6
+ "_name_or_path": "",
7
+ "architectures": null,
8
+ "audio_embed_dim": 640,
9
+ "chunk_size_feed_forward": 0,
10
+ "dtype": "bfloat16",
11
+ "id2label": {
12
+ "0": "LABEL_0",
13
+ "1": "LABEL_1"
14
+ },
15
+ "initializer_range": 0.02,
16
+ "is_encoder_decoder": false,
17
+ "label2id": {
18
+ "LABEL_0": 0,
19
+ "LABEL_1": 1
20
+ },
21
+ "model_type": "gemma4_unified_audio",
22
+ "output_attentions": false,
23
+ "output_hidden_states": false,
24
+ "problem_type": null,
25
+ "return_dict": true,
26
+ "rms_norm_eps": 1e-06
27
+ },
28
+ "audio_token_id": 258881,
29
+ "boa_token_id": 256000,
30
+ "boi_token_id": 255999,
31
+ "bos_token_id": 2,
32
+ "dtype": "bfloat16",
33
+ "eoa_token_index": 258883,
34
+ "eoi_token_id": 258882,
35
+ "eos_token_id": 1,
36
+ "hidden_size": 3840,
37
+ "image_token_id": 258880,
38
+ "initializer_range": 0.02,
39
+ "model_type": "gemma4_unified",
40
+ "pad_token_id": 0,
41
+ "text_config": {
42
+ "attention_bias": false,
43
+ "attention_dropout": 0.0,
44
+ "attention_k_eq_v": true,
45
+ "bos_token_id": 2,
46
+ "dtype": "bfloat16",
47
+ "enable_moe_block": false,
48
+ "eos_token_id": 1,
49
+ "final_logit_softcapping": 30.0,
50
+ "global_head_dim": 512,
51
+ "head_dim": 256,
52
+ "hidden_activation": "gelu_pytorch_tanh",
53
+ "hidden_size": 3840,
54
+ "hidden_size_per_layer_input": 0,
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 15360,
57
+ "layer_types": [
58
+ "sliding_attention",
59
+ "sliding_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "full_attention",
64
+ "sliding_attention",
65
+ "sliding_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "full_attention",
70
+ "sliding_attention",
71
+ "sliding_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "full_attention",
76
+ "sliding_attention",
77
+ "sliding_attention",
78
+ "sliding_attention",
79
+ "sliding_attention",
80
+ "sliding_attention",
81
+ "full_attention",
82
+ "sliding_attention",
83
+ "sliding_attention",
84
+ "sliding_attention",
85
+ "sliding_attention",
86
+ "sliding_attention",
87
+ "full_attention",
88
+ "sliding_attention",
89
+ "sliding_attention",
90
+ "sliding_attention",
91
+ "sliding_attention",
92
+ "sliding_attention",
93
+ "full_attention",
94
+ "sliding_attention",
95
+ "sliding_attention",
96
+ "sliding_attention",
97
+ "sliding_attention",
98
+ "sliding_attention",
99
+ "full_attention",
100
+ "sliding_attention",
101
+ "sliding_attention",
102
+ "sliding_attention",
103
+ "sliding_attention",
104
+ "sliding_attention",
105
+ "full_attention"
106
+ ],
107
+ "max_position_embeddings": 262144,
108
+ "model_type": "gemma4_unified_text",
109
+ "moe_intermediate_size": null,
110
+ "num_attention_heads": 16,
111
+ "num_experts": null,
112
+ "num_global_key_value_heads": 1,
113
+ "num_hidden_layers": 48,
114
+ "num_key_value_heads": 8,
115
+ "num_kv_shared_layers": 0,
116
+ "pad_token_id": 0,
117
+ "rms_norm_eps": 1e-06,
118
+ "rope_parameters": {
119
+ "full_attention": {
120
+ "partial_rotary_factor": 0.25,
121
+ "rope_theta": 1000000.0,
122
+ "rope_type": "proportional"
123
+ },
124
+ "sliding_attention": {
125
+ "rope_theta": 10000.0,
126
+ "rope_type": "default"
127
+ }
128
+ },
129
+ "sliding_window": 1024,
130
+ "tie_word_embeddings": true,
131
+ "top_k_experts": null,
132
+ "use_bidirectional_attention": "vision",
133
+ "use_cache": false,
134
+ "use_double_wide_mlp": false,
135
+ "vocab_size": 262144,
136
+ "vocab_size_per_layer_input": 262144
137
+ },
138
+ "tie_word_embeddings": true,
139
+ "transformers_version": "5.10.0",
140
+ "use_cache": false,
141
+ "video_token_id": 258884,
142
+ "vision_config": {
143
+ "_name_or_path": "",
144
+ "architectures": null,
145
+ "chunk_size_feed_forward": 0,
146
+ "dtype": "bfloat16",
147
+ "id2label": {
148
+ "0": "LABEL_0",
149
+ "1": "LABEL_1"
150
+ },
151
+ "initializer_range": 0.02,
152
+ "is_encoder_decoder": false,
153
+ "label2id": {
154
+ "LABEL_0": 0,
155
+ "LABEL_1": 1
156
+ },
157
+ "mm_embed_dim": 3840,
158
+ "mm_posemb_size": 1120,
159
+ "model_type": "gemma4_unified_vision",
160
+ "num_soft_tokens": 280,
161
+ "output_attentions": false,
162
+ "output_hidden_states": false,
163
+ "output_proj_dims": 3840,
164
+ "patch_size": 16,
165
+ "pooling_kernel_size": 3,
166
+ "problem_type": null,
167
+ "return_dict": true,
168
+ "rms_norm_eps": 1e-06
169
+ }
170
+ }
200M/generation_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 1,
6
+ 1
7
+ ],
8
+ "pad_token_id": 0,
9
+ "suppress_tokens": [
10
+ 258883,
11
+ 258882
12
+ ],
13
+ "temperature": 1.0,
14
+ "top_k": 64,
15
+ "top_p": 0.95,
16
+ "transformers_version": "5.10.0"
17
+ }
200M/latest ADDED
@@ -0,0 +1 @@
 
 
1
+ global_step2035
200M/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29b8b4e32fcd4506d97631e4937f869223a7e0b3bad6da37de8501f0d1a5672b
3
+ size 23919549408
200M/processor_config.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audio_ms_per_token": 40,
3
+ "audio_seq_length": 750,
4
+ "feature_extractor": {
5
+ "audio_samples_per_token": 640,
6
+ "feature_extractor_type": "Gemma4UnifiedAudioFeatureExtractor",
7
+ "feature_size": 640,
8
+ "padding_side": "right",
9
+ "padding_value": 0.0,
10
+ "return_attention_mask": true,
11
+ "sampling_rate": 16000
12
+ },
13
+ "image_processor": {
14
+ "do_convert_rgb": true,
15
+ "do_normalize": false,
16
+ "do_rescale": true,
17
+ "do_resize": true,
18
+ "image_mean": [
19
+ 0.0,
20
+ 0.0,
21
+ 0.0
22
+ ],
23
+ "image_processor_type": "Gemma4UnifiedImageProcessor",
24
+ "image_std": [
25
+ 1.0,
26
+ 1.0,
27
+ 1.0
28
+ ],
29
+ "max_soft_tokens": 280,
30
+ "patch_size": 16,
31
+ "pooling_kernel_size": 3,
32
+ "resample": 3,
33
+ "rescale_factor": 0.00392156862745098
34
+ },
35
+ "image_seq_length": 280,
36
+ "processor_class": "Gemma4UnifiedProcessor",
37
+ "video_processor": {
38
+ "do_convert_rgb": true,
39
+ "do_normalize": true,
40
+ "do_rescale": true,
41
+ "do_resize": true,
42
+ "do_sample_frames": true,
43
+ "image_mean": [
44
+ 0.0,
45
+ 0.0,
46
+ 0.0
47
+ ],
48
+ "image_std": [
49
+ 1.0,
50
+ 1.0,
51
+ 1.0
52
+ ],
53
+ "max_soft_tokens": 70,
54
+ "num_frames": 32,
55
+ "patch_size": 16,
56
+ "pooling_kernel_size": 3,
57
+ "resample": 3,
58
+ "rescale_factor": 0.00392156862745098,
59
+ "return_metadata": false,
60
+ "video_processor_type": "Gemma4UnifiedVideoProcessor"
61
+ }
62
+ }
200M/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d7dcf54a8dbd1fff243640ddc034149fbe7c90ed70b8a95a6bb093e989a7577
3
+ size 14469
200M/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2a2b0b9d330140afd742a57df45106276d425d38a78e53b0fd43dba9559f8d4
3
+ size 14469
200M/rng_state_10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ce222f8f11fd62fd6cb82264500aefbc9e7ef16907ed3e4aa3a8a99c21d95d3
3
+ size 14476
200M/rng_state_11.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d07144bfc4ed675dda70ab66923d17d6f9233560b2db09689cbe1f262c590c88
3
+ size 14476
200M/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5db634f6ed5151d3c898fbe3c365f950516c404bed926de16b253b26e63c7cf
3
+ size 14469
200M/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad774eade44525098fff2d5910e3d16565fca4c2af29ed0bc4d9315a8ac9607a
3
+ size 14469
200M/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebdee1b37be2238764bcfeed9dad268e938248152b0fba60eaf89adda99eac0f
3
+ size 14469
200M/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6046157bd3d69a7cf6ee508b7359b30cb71488a11b11b07a614041b06813c553
3
+ size 14469
200M/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebc9300f101a324d0361f01aec1d57e26b57a703b4e536f4fe982c8ba57a9144
3
+ size 14469
200M/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:357df7b609d6ffd7f22b998281f7856a711ee83505e5ed2835959554b0185df7
3
+ size 14469
200M/rng_state_8.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fb33667d752a94ade6841f62b8b8fcc22256b4712bd568b141bc690d4a42e11
3
+ size 14469
200M/rng_state_9.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:354a7a38e97ba11b8ad29bf4d00d74f92295b32124cff0ef46baa862aa9b0f2f
3
+ size 14469
200M/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea084231345496123a1e2e3e6e6377891838cfabbf1b42517ab7575e5966916e
3
+ size 1465
200M/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12bac982b793c44b03d52a250a9f0d0b666813da566b910c24a6da0695fd11e6
3
+ size 32170070
200M/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audio_token": "<|audio|>",
3
+ "backend": "tokenizers",
4
+ "boa_token": "<|audio>",
5
+ "boi_token": "<|image>",
6
+ "bos_token": "<bos>",
7
+ "eoa_token": "<audio|>",
8
+ "eoc_token": "<channel|>",
9
+ "eoi_token": "<image|>",
10
+ "eos_token": "<eos>",
11
+ "eot_token": "<turn|>",
12
+ "escape_token": "<|\"|>",
13
+ "etc_token": "<tool_call|>",
14
+ "etd_token": "<tool|>",
15
+ "etr_token": "<tool_response|>",
16
+ "extra_special_tokens": [
17
+ "<|video|>"
18
+ ],
19
+ "image_token": "<|image|>",
20
+ "is_local": true,
21
+ "local_files_only": true,
22
+ "mask_token": "<mask>",
23
+ "model_max_length": 1000000000000000019884624838656,
24
+ "model_specific_special_tokens": {
25
+ "audio_token": "<|audio|>",
26
+ "boa_token": "<|audio>",
27
+ "boi_token": "<|image>",
28
+ "eoa_token": "<audio|>",
29
+ "eoc_token": "<channel|>",
30
+ "eoi_token": "<image|>",
31
+ "eot_token": "<turn|>",
32
+ "escape_token": "<|\"|>",
33
+ "etc_token": "<tool_call|>",
34
+ "etd_token": "<tool|>",
35
+ "etr_token": "<tool_response|>",
36
+ "image_token": "<|image|>",
37
+ "soc_token": "<|channel>",
38
+ "sot_token": "<|turn>",
39
+ "stc_token": "<|tool_call>",
40
+ "std_token": "<|tool>",
41
+ "str_token": "<|tool_response>",
42
+ "think_token": "<|think|>"
43
+ },
44
+ "pad_token": "<pad>",
45
+ "padding_side": "right",
46
+ "processor_class": "Gemma4UnifiedProcessor",
47
+ "soc_token": "<|channel>",
48
+ "sot_token": "<|turn>",
49
+ "split_special_tokens": false,
50
+ "stc_token": "<|tool_call>",
51
+ "std_token": "<|tool>",
52
+ "str_token": "<|tool_response>",
53
+ "think_token": "<|think|>",
54
+ "tokenizer_class": "GemmaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
200M/trainer_state.json ADDED
@@ -0,0 +1,1455 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.05579775712209701,
6
+ "eval_steps": 500,
7
+ "global_step": 2035,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0002741904526884374,
14
+ "grad_norm": 533.0549778145878,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 2.0392507553100585,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0005483809053768748,
21
+ "grad_norm": 129.62802977018143,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 1.1088291168212892,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.0008225713580653121,
28
+ "grad_norm": 37.028700378797375,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.8763736724853516,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.0010967618107537496,
35
+ "grad_norm": 11.415105214785655,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.8334451675415039,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.0013709522634421868,
42
+ "grad_norm": 168.0092220161655,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.841560173034668,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.0016451427161306242,
49
+ "grad_norm": 10.9544582922062,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.7757131099700928,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0019193331688190617,
56
+ "grad_norm": 52.12017293008163,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.7379593849182129,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.0021935236215074993,
63
+ "grad_norm": 24.23368291019739,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.871122169494629,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.0024677140741959365,
70
+ "grad_norm": 44.606225719656116,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.8103988647460938,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.0027419045268843737,
77
+ "grad_norm": 19.394878252919106,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.8088878631591797,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.0030160949795728113,
84
+ "grad_norm": 154.65284754287825,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.7940959930419922,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.0032902854322612485,
91
+ "grad_norm": 58.453168158560985,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.8113164901733398,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.003564475884949686,
98
+ "grad_norm": 121.71906195134056,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.802008056640625,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.0038386663376381233,
105
+ "grad_norm": 7.063904217729592,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 0.7886086940765381,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.004112856790326561,
112
+ "grad_norm": 5.886231107848548,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.8553682327270508,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.004387047243014999,
119
+ "grad_norm": 17.479053516912124,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.878786849975586,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.004661237695703435,
126
+ "grad_norm": 15.809884339669948,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.8892171859741211,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.004935428148391873,
133
+ "grad_norm": 15.552804520715554,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.907989501953125,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.005209618601080311,
140
+ "grad_norm": 8.234516558585694,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.8959874153137207,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.005483809053768747,
147
+ "grad_norm": 47.07479845477993,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.7972547054290772,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.005757999506457185,
154
+ "grad_norm": 14.678891090428264,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.7945639610290527,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.006032189959145623,
161
+ "grad_norm": 17.99623194706809,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.8467901229858399,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.00630638041183406,
168
+ "grad_norm": 51.95652459415379,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.7964937210083007,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.006580570864522497,
175
+ "grad_norm": 10.738865490640423,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.7535751342773438,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.006854761317210935,
182
+ "grad_norm": 10.795608734038868,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.828792953491211,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.007128951769899372,
189
+ "grad_norm": 14.551689425739292,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.8123558044433594,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.00740314222258781,
196
+ "grad_norm": 26.532992867443514,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.8517896652221679,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.007677332675276247,
203
+ "grad_norm": 34.07990246963607,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.8389625549316406,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.007951523127964684,
210
+ "grad_norm": 27.19116883891022,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.8190082550048828,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.008225713580653122,
217
+ "grad_norm": 16.718638113526183,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.8008914947509765,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.00849990403334156,
224
+ "grad_norm": 27.009662680205103,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.8091920852661133,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.008774094486029997,
231
+ "grad_norm": 9.031648650732553,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.8081918716430664,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.009048284938718433,
238
+ "grad_norm": 15.927828911998665,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.7878878593444825,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.00932247539140687,
245
+ "grad_norm": 6.019580282287438,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.8092373847961426,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.009596665844095308,
252
+ "grad_norm": 10.097656044283081,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.8236124038696289,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.009870856296783746,
259
+ "grad_norm": 3.898510274122292,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.8201662063598633,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.010145046749472184,
266
+ "grad_norm": 10.433240846124043,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.7954160213470459,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.010419237202160621,
273
+ "grad_norm": 3.9109175859167626,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.8451309204101562,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.010693427654849059,
280
+ "grad_norm": 19.47785359214917,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.8227203369140625,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.010967618107537495,
287
+ "grad_norm": 1.9979422322935754,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.7947024345397949,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.011241808560225932,
294
+ "grad_norm": 2.1152537387703063,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.7340120315551758,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.01151599901291437,
301
+ "grad_norm": 2.457083494187141,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.7300191879272461,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.011790189465602808,
308
+ "grad_norm": 17.457321034227927,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.7896897315979003,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.012064379918291245,
315
+ "grad_norm": 3.7757446450653047,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.7420886039733887,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.012338570370979683,
322
+ "grad_norm": 10.718659609403895,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.7464280605316163,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.01261276082366812,
329
+ "grad_norm": 13.096467752661693,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.7420991420745849,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.012886951276356558,
336
+ "grad_norm": 7.554114077938612,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.7896678924560547,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.013161141729044994,
343
+ "grad_norm": 1.9610155148576782,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.7641692161560059,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.013435332181733432,
350
+ "grad_norm": 3.743726960430974,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.7724472999572753,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.01370952263442187,
357
+ "grad_norm": 3.6205794374072897,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.7726649284362793,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.013983713087110307,
364
+ "grad_norm": 2.751894400685354,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.7311810970306396,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.014257903539798744,
371
+ "grad_norm": 15.8354606407861,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.7479830741882324,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.014532093992487182,
378
+ "grad_norm": 9.741291005928273,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.758505916595459,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.01480628444517562,
385
+ "grad_norm": 6.361558235058739,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.7497701644897461,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.015080474897864056,
392
+ "grad_norm": 12.406630096904701,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.7578043937683105,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.015354665350552493,
399
+ "grad_norm": 5.437533111150059,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.7494767665863037,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.015628855803240933,
406
+ "grad_norm": 5.385414434765143,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.76002197265625,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.01590304625592937,
413
+ "grad_norm": 2.514872003150109,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.753857946395874,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.016177236708617804,
420
+ "grad_norm": 7.350990110651167,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.7598513603210449,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.016451427161306244,
427
+ "grad_norm": 9.359850474100211,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.7978845596313476,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.01672561761399468,
434
+ "grad_norm": 2.5927971608846243,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.7353355407714843,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.01699980806668312,
441
+ "grad_norm": 13.580441280433185,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.7530646800994873,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.017273998519371555,
448
+ "grad_norm": 7.032504711444772,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.7999831199645996,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.017548188972059994,
455
+ "grad_norm": 3.7623518166414134,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.7583576202392578,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.01782237942474843,
462
+ "grad_norm": 9.996121547764488,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.7126437187194824,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.018096569877436866,
469
+ "grad_norm": 3.8005663709948423,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.7714213371276856,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.018370760330125305,
476
+ "grad_norm": 2.2803953101797445,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.735899829864502,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.01864495078281374,
483
+ "grad_norm": 2.2804717033812687,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.7463047981262207,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.01891914123550218,
490
+ "grad_norm": 2.6630744686256174,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.7112466812133789,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.019193331688190617,
497
+ "grad_norm": 8.59421748677637,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.8032501220703125,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.019467522140879056,
504
+ "grad_norm": 6.839099911030037,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.7257285118103027,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.019741712593567492,
511
+ "grad_norm": 6.527196880605798,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.7394065856933594,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.020015903046255928,
518
+ "grad_norm": 4.91546686150854,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.7233449459075928,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.020290093498944367,
525
+ "grad_norm": 9.615722695591046,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.714713191986084,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.020564283951632803,
532
+ "grad_norm": 2.6469047522209244,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.7250522613525391,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.020838474404321242,
539
+ "grad_norm": 27.513230581702402,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.709619665145874,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.021112664857009678,
546
+ "grad_norm": 2.3305572520941102,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.7650126457214356,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.021386855309698118,
553
+ "grad_norm": 1.7645473455518579,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.7501170635223389,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.021661045762386553,
560
+ "grad_norm": 2.706703322701269,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.7390304565429687,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.02193523621507499,
567
+ "grad_norm": 1.343452945442269,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.7560922145843506,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.02220942666776343,
574
+ "grad_norm": 1.8131674161495455,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.7617329597473145,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.022483617120451865,
581
+ "grad_norm": 6.6257304148619,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.7030315399169922,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.022757807573140304,
588
+ "grad_norm": 2.7477972541714237,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.7051456451416016,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.02303199802582874,
595
+ "grad_norm": 1.8551689554260649,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.790980863571167,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.02330618847851718,
602
+ "grad_norm": 4.091035715965749,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.7167482376098633,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.023580378931205615,
609
+ "grad_norm": 3.13444342764459,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.7428287506103516,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.023854569383894055,
616
+ "grad_norm": 7.428502868185567,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.7720952510833741,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.02412875983658249,
623
+ "grad_norm": 2.738880412295684,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.7307533264160156,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.024402950289270926,
630
+ "grad_norm": 5.442319700482665,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.7590367794036865,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.024677140741959366,
637
+ "grad_norm": 4.209775631505361,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.7398377418518066,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.0249513311946478,
644
+ "grad_norm": 3.317366374639838,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.7573592662811279,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.02522552164733624,
651
+ "grad_norm": 2.21708523073645,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.7442788124084473,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.025499712100024677,
658
+ "grad_norm": 5.003510596258695,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.7208542823791504,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.025773902552713116,
665
+ "grad_norm": 1.8666696979850983,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.7219260215759278,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.026048093005401552,
672
+ "grad_norm": 2.319707094464392,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.705035400390625,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.026322283458089988,
679
+ "grad_norm": 1.559433923602148,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.7253781795501709,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.026596473910778427,
686
+ "grad_norm": 3.1103622605289294,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.7434506416320801,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.026870664363466863,
693
+ "grad_norm": 4.482952591971397,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.7423406600952148,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.027144854816155303,
700
+ "grad_norm": 2.349643537640656,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.7380632400512696,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.02741904526884374,
707
+ "grad_norm": 6.624084473754211,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.7448741912841796,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.027693235721532178,
714
+ "grad_norm": 1.959542744744174,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.7620717048645019,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.027967426174220614,
721
+ "grad_norm": 1.9359370487542638,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.7283543586730957,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.02824161662690905,
728
+ "grad_norm": 3.7510091510273424,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.7341510772705078,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.02851580707959749,
735
+ "grad_norm": 1.645233108914731,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.7321323394775391,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.028789997532285925,
742
+ "grad_norm": 5.060129114701537,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.7923129558563232,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.029064187984974364,
749
+ "grad_norm": 6.767343077930711,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.7364850997924804,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.0293383784376628,
756
+ "grad_norm": 7.157354259172915,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.7175948143005371,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.02961256889035124,
763
+ "grad_norm": 10.118526445754803,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.6956705093383789,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.029886759343039675,
770
+ "grad_norm": 3.2975128574851205,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.7435084342956543,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.03016094979572811,
777
+ "grad_norm": 3.79438108252297,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.7209542274475098,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.03043514024841655,
784
+ "grad_norm": 3.6413783178286896,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.7462982177734375,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.030709330701104987,
791
+ "grad_norm": 2.4359365603997323,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.7772771835327148,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.030983521153793426,
798
+ "grad_norm": 2.067253539210089,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.754563283920288,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.031257711606481865,
805
+ "grad_norm": 1.3603842511083,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.7163387298583984,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.0315319020591703,
812
+ "grad_norm": 7.634204958345298,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.7153648376464844,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.03180609251185874,
819
+ "grad_norm": 2.224287938133906,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.7743209838867188,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.03208028296454717,
826
+ "grad_norm": 8.328106347948369,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.7125131130218506,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.03235447341723561,
833
+ "grad_norm": 4.9140382117559716,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.734964656829834,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.03262866386992405,
840
+ "grad_norm": 1.6437427011396712,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.736287260055542,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.03290285432261249,
847
+ "grad_norm": 1.7984413046378422,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.6888255119323731,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.03317704477530092,
854
+ "grad_norm": 3.003273429621072,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.7354516506195068,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.03345123522798936,
861
+ "grad_norm": 1.581766516298146,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.7175889492034913,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.0337254256806778,
868
+ "grad_norm": 2.3247470513343185,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.714189863204956,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.03399961613336624,
875
+ "grad_norm": 7.9334494235192325,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.7136695861816407,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.034273806586054674,
882
+ "grad_norm": 49.62487496955337,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.7086945533752441,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.03454799703874311,
889
+ "grad_norm": 2.890774584879648,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.7269087791442871,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.034822187491431546,
896
+ "grad_norm": 4.796531005737684,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.7484929084777832,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.03509637794411999,
903
+ "grad_norm": 1.6372708148105157,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.6999114036560059,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.035370568396808424,
910
+ "grad_norm": 4.162341191850801,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.7129648208618165,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.03564475884949686,
917
+ "grad_norm": 4.017056556093628,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.7217267036437989,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.035918949302185296,
924
+ "grad_norm": 9.786655951652824,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.7361152648925782,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.03619313975487373,
931
+ "grad_norm": 17.798888728639277,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.7423253059387207,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.036467330207562175,
938
+ "grad_norm": 25.98230175781132,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.7079302787780761,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.03674152066025061,
945
+ "grad_norm": 4.107909151191186,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.7637813568115235,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.03701571111293905,
952
+ "grad_norm": 1.8177555687046802,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.7440331459045411,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.03728990156562748,
959
+ "grad_norm": 6.631474920460703,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.7420357704162598,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.037564092018315925,
966
+ "grad_norm": 1.182591249805126,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.7243555068969727,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.03783828247100436,
973
+ "grad_norm": 1.5239290367740697,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.7780046463012695,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.0381124729236928,
980
+ "grad_norm": 1.9611702789468477,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.7606742858886719,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.03838666337638123,
987
+ "grad_norm": 3.4538398959809564,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.7161018371582031,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.03866085382906967,
994
+ "grad_norm": 6.74405592418129,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.7269341945648193,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.03893504428175811,
1001
+ "grad_norm": 5.0613023509206,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.7150623321533203,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.03920923473444655,
1008
+ "grad_norm": 15.935601654785444,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.6991084098815918,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.039483425187134984,
1015
+ "grad_norm": 189.58206876056943,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.715428352355957,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.03975761563982342,
1022
+ "grad_norm": 13.224258279776146,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.7448666572570801,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.040031806092511855,
1029
+ "grad_norm": 15.23890210961455,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.7545663833618164,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.0403059965452003,
1036
+ "grad_norm": 2.089772579003186,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.7143968105316162,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.040580186997888734,
1043
+ "grad_norm": 672.5570249806746,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.6955307960510254,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.04085437745057717,
1050
+ "grad_norm": 2.499047720535622,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.7183216094970704,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.041128567903265606,
1057
+ "grad_norm": 1.6820671266931866,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.7442133903503418,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.04140275835595405,
1064
+ "grad_norm": 1.6322487359391142,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.7415005683898925,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.041676948808642485,
1071
+ "grad_norm": 2.0768691766562974,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.7350949764251709,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.04195113926133092,
1078
+ "grad_norm": 3.462638107939682,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.6958516597747803,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.042225329714019356,
1085
+ "grad_norm": 1.9410342316482636,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.7019001960754394,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.04249952016670779,
1092
+ "grad_norm": 2.2435234267896944,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.7038233757019043,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.042773710619396235,
1099
+ "grad_norm": 50.31729516447788,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.6971765041351319,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.04304790107208467,
1106
+ "grad_norm": 6.726153221416043,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.7901406764984131,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.04332209152477311,
1113
+ "grad_norm": 25.290061958502374,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.7404637813568116,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.04359628197746154,
1120
+ "grad_norm": 11.648387859593914,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.7325190544128418,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.04387047243014998,
1127
+ "grad_norm": 3.998377511327454,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.6938001155853272,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.04414466288283842,
1134
+ "grad_norm": 1.4609573707991412,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.6974159240722656,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.04441885333552686,
1141
+ "grad_norm": 1.2908767847499276,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.735476303100586,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.04469304378821529,
1148
+ "grad_norm": 1.5263587301999937,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.6849418640136719,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.04496723424090373,
1155
+ "grad_norm": 2.71012026891521,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.6952274322509766,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.04524142469359217,
1162
+ "grad_norm": 1.5738552850120824,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.7284816265106201,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.04551561514628061,
1169
+ "grad_norm": 1.5435257666345013,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.7115949153900146,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.045789805598969044,
1176
+ "grad_norm": 1.516425220632835,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.6966533660888672,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.04606399605165748,
1183
+ "grad_norm": 1.4858677693562772,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.712862491607666,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.046338186504345916,
1190
+ "grad_norm": 1.1928774431737874,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.715973949432373,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.04661237695703436,
1197
+ "grad_norm": 1.3801750667103458,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.6785056114196777,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.046886567409722794,
1204
+ "grad_norm": 1.1562227123885702,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.6806015014648438,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.04716075786241123,
1211
+ "grad_norm": 1.966065169897555,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.6789681911468506,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.047434948315099666,
1218
+ "grad_norm": 1.796240102087905,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.6643640041351319,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.04770913876778811,
1225
+ "grad_norm": 1.4714599948408273,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.672383975982666,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.047983329220476545,
1232
+ "grad_norm": 1.05613596470809,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.6921384334564209,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.04825751967316498,
1239
+ "grad_norm": 1.168321764739896,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.6853407859802246,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.04853171012585342,
1246
+ "grad_norm": 1.304231611987027,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.6831422328948975,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.04880590057854185,
1253
+ "grad_norm": 1.271885101891083,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.7312606811523438,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.049080091031230295,
1260
+ "grad_norm": 1.1274168419233672,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.7195809841156006,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.04935428148391873,
1267
+ "grad_norm": 2.1060356897000196,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.6862109184265137,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.04962847193660717,
1274
+ "grad_norm": 1.1668574261874323,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.7068725109100342,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.0499026623892956,
1281
+ "grad_norm": 3.1583477159166673,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.7150220394134521,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.05017685284198404,
1288
+ "grad_norm": 1.5843620172233683,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.66549072265625,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.05045104329467248,
1295
+ "grad_norm": 1.2547928536390052,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.7375857353210449,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.05072523374736092,
1302
+ "grad_norm": 4.29725259279178,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.6876885890960693,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.050999424200049354,
1309
+ "grad_norm": 1.9025925412069744,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.7387830734252929,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.05127361465273779,
1316
+ "grad_norm": 1.5406949311604028,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.6733179092407227,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.05154780510542623,
1323
+ "grad_norm": 20.215212215144863,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.7654403686523438,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.05182199555811467,
1330
+ "grad_norm": 1.5669531048405803,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.7350390911102295,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.052096186010803104,
1337
+ "grad_norm": 7.2191546562371425,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.7414397716522216,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.05237037646349154,
1344
+ "grad_norm": 1.2325274169727398,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.7655929088592529,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.052644566916179976,
1351
+ "grad_norm": 1.1944179026527708,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.7075240135192871,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.05291875736886842,
1358
+ "grad_norm": 1.302822879313398,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.6797014236450195,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.053192947821556855,
1365
+ "grad_norm": 2.0168747405099747,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.7000571727752686,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.05346713827424529,
1372
+ "grad_norm": 1.6344897385611805,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.7235137939453125,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.053741328726933726,
1379
+ "grad_norm": 3.5766753006414125,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.7371991634368896,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.05401551917962216,
1386
+ "grad_norm": 1.4953479982574733,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.7229987621307373,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.054289709632310605,
1393
+ "grad_norm": 1.6598347475638233,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.6758342742919922,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.05456390008499904,
1400
+ "grad_norm": 1.3876361065240026,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.6621166229248047,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.05483809053768748,
1407
+ "grad_norm": 1.2953540155757288,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.701081657409668,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.05511228099037591,
1414
+ "grad_norm": 1.0757376098134102,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.6949031829833985,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.055386471443064356,
1421
+ "grad_norm": 1.3515745918409625,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.6849285125732422,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.05566066189575279,
1428
+ "grad_norm": 3.4678908742290884,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.6919690608978272,
1431
+ "step": 2030
1432
+ }
1433
+ ],
1434
+ "logging_steps": 10,
1435
+ "max_steps": 5935,
1436
+ "num_input_tokens_seen": 0,
1437
+ "num_train_epochs": 1,
1438
+ "save_steps": 2035,
1439
+ "stateful_callbacks": {
1440
+ "TrainerControl": {
1441
+ "args": {
1442
+ "should_epoch_stop": false,
1443
+ "should_evaluate": false,
1444
+ "should_log": false,
1445
+ "should_save": true,
1446
+ "should_training_stop": false
1447
+ },
1448
+ "attributes": {}
1449
+ }
1450
+ },
1451
+ "total_flos": 963613493821440.0,
1452
+ "train_batch_size": 1,
1453
+ "trial_name": null,
1454
+ "trial_params": null
1455
+ }
200M/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:392ea69d123a584ac5b993c15219b31093e4d98dddc8f0c3122fc776ee8c9c55
3
+ size 7313