Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-g2_4_plan_cm_oldcode_newcode

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
db16f52
·
verified ·
1 Parent(s): 55193ff

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:e4b833cedfaabbd4891ebf4cae307b5c523ac9b910ee971a91bb88e555a325b4
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": 0.06519719347707686,
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": 0.0003203793291256848,
14
+ "grad_norm": 2697.9695951365356,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 2.1080955505371093,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0006407586582513696,
21
+ "grad_norm": 150.28792394404246,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 1.051735782623291,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.0009611379873770544,
28
+ "grad_norm": 11.728004906904669,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.8452852249145508,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.0012815173165027392,
35
+ "grad_norm": 79.82045143025138,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.8374807357788085,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.001601896645628424,
42
+ "grad_norm": 15.487828868814473,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.8031275749206543,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.0019222759747541088,
49
+ "grad_norm": 26.189373840069653,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.7619327545166016,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.002242655303879794,
56
+ "grad_norm": 24.510054444181353,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.788119888305664,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.0025630346330054784,
63
+ "grad_norm": 27.619677813912134,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.8074554443359375,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.0028834139621311634,
70
+ "grad_norm": 53.27561070809039,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.7780509948730469,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.003203793291256848,
77
+ "grad_norm": 20.22293994140261,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.7639347076416015,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.003524172620382533,
84
+ "grad_norm": 28.55291130912943,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.7700469970703125,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.0038445519495082176,
91
+ "grad_norm": 7.520305767933223,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.7953043937683105,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.004164931278633903,
98
+ "grad_norm": 96.08078842951305,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.8506025314331055,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.004485310607759588,
105
+ "grad_norm": 948.131373093307,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 0.8122509002685547,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.004805689936885272,
112
+ "grad_norm": 128.55037969805156,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.7938695907592773,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.005126069266010957,
119
+ "grad_norm": 25.93618158094234,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.8438175201416016,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.005446448595136642,
126
+ "grad_norm": 9.528266971779534,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.9011789321899414,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.005766827924262327,
133
+ "grad_norm": 4.6353029582495315,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.8190475463867187,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.006087207253388012,
140
+ "grad_norm": 7.804600561994388,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.7879319667816163,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.006407586582513696,
147
+ "grad_norm": 8.759193053432988,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.8373113632202148,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.006727965911639381,
154
+ "grad_norm": 5.7645630188571495,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.8225456237792969,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.007048345240765066,
161
+ "grad_norm": 2.975476645382291,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.7898527145385742,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.007368724569890751,
168
+ "grad_norm": 3.5660713690110923,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.7848415374755859,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.007689103899016435,
175
+ "grad_norm": 3.776542520178389,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.7830493927001954,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.008009483228142121,
182
+ "grad_norm": 5.652886706438173,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.7420679092407226,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.008329862557267805,
189
+ "grad_norm": 2.7433625896377887,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.8062055587768555,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.00865024188639349,
196
+ "grad_norm": 4.993469336501579,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.7772369384765625,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.008970621215519175,
203
+ "grad_norm": 5.978072597304008,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.7567183494567871,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.00929100054464486,
210
+ "grad_norm": 20.0019916771175,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.7632350921630859,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.009611379873770544,
217
+ "grad_norm": 1813.803400221406,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.7820405006408692,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.00993175920289623,
224
+ "grad_norm": 15.257478921571794,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.8594757080078125,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.010252138532021914,
231
+ "grad_norm": 26.97548621262987,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.8610923767089844,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.0105725178611476,
238
+ "grad_norm": 21.892790276243126,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.8686389923095703,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.010892897190273284,
245
+ "grad_norm": 16.06260184223094,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.8639827728271484,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.011213276519398968,
252
+ "grad_norm": 29.260614803206966,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.7904685497283935,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.011533655848524654,
259
+ "grad_norm": 16.44688311222318,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.8587644577026368,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.011854035177650338,
266
+ "grad_norm": 70.37237245987374,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.8287432670593262,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.012174414506776024,
273
+ "grad_norm": 7.352519778805397,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.8243340492248535,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.012494793835901708,
280
+ "grad_norm": 7.57180002114558,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.8488252639770508,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.012815173165027392,
287
+ "grad_norm": 4.028963614155682,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.800967788696289,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.013135552494153078,
294
+ "grad_norm": 7.604607607710578,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.8435401916503906,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.013455931823278762,
301
+ "grad_norm": 5.686576942826079,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.8086394309997559,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.013776311152404446,
308
+ "grad_norm": 29.16934098168494,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.8092558860778809,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.014096690481530132,
315
+ "grad_norm": 7.838527647928418,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.7873313903808594,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.014417069810655816,
322
+ "grad_norm": 4.360151373268049,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.8105262756347656,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.014737449139781502,
329
+ "grad_norm": 3.093592672180748,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.7954371452331543,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.015057828468907186,
336
+ "grad_norm": 2.8867780320165948,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.7867568969726563,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.01537820779803287,
343
+ "grad_norm": 2.5525945956595404,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.7605972290039062,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.015698587127158554,
350
+ "grad_norm": 4.350350397516954,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.7946616172790527,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.016018966456284242,
357
+ "grad_norm": 2.5601326184252624,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.7657347679138183,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.016339345785409926,
364
+ "grad_norm": 3.0411596315285943,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.7469792366027832,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.01665972511453561,
371
+ "grad_norm": 1.6677799044175043,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.737004280090332,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.016980104443661295,
378
+ "grad_norm": 4.326231014259539,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.7799790859222412,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.01730048377278698,
385
+ "grad_norm": 2.732975416289203,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.8164985656738282,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.017620863101912666,
392
+ "grad_norm": 2.1962459439491036,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.7743217468261718,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.01794124243103835,
399
+ "grad_norm": 4.329667095154606,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.8150390625,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.018261621760164035,
406
+ "grad_norm": 1.7505089770107007,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.8161712646484375,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.01858200108928972,
413
+ "grad_norm": 1.94080520422082,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.7340856075286866,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.018902380418415403,
420
+ "grad_norm": 3.767043072114596,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.7465179443359375,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.019222759747541087,
427
+ "grad_norm": 2.134739515696804,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.7511123657226563,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.019543139076666775,
434
+ "grad_norm": 2.6553715328436724,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.7474520683288575,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.01986351840579246,
441
+ "grad_norm": 23.554705629944248,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.7547467708587646,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.020183897734918143,
448
+ "grad_norm": 9.663715178820514,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.7591798782348633,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.020504277064043827,
455
+ "grad_norm": 2.857655614152204,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.7372615814208985,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.02082465639316951,
462
+ "grad_norm": 19.794443617215876,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.7473020076751709,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.0211450357222952,
469
+ "grad_norm": 3.4519708452708513,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.7669941902160644,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.021465415051420883,
476
+ "grad_norm": 4.439475537657204,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.7708707809448242,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.021785794380546567,
483
+ "grad_norm": 4.288236090232602,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.75105299949646,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.02210617370967225,
490
+ "grad_norm": 19.247483032059645,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.7456221580505371,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.022426553038797935,
497
+ "grad_norm": 6.277608344095431,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.772921371459961,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.022746932367923623,
504
+ "grad_norm": 3.5768536610883284,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.7416125297546386,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.023067311697049307,
511
+ "grad_norm": 15.05269132810213,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.7103744506835937,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.02338769102617499,
518
+ "grad_norm": 4.468376174246059,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.727785873413086,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.023708070355300676,
525
+ "grad_norm": 5.087071299517523,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.7700984477996826,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.02402844968442636,
532
+ "grad_norm": 11.288253253171709,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.7581868648529053,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.024348829013552047,
539
+ "grad_norm": 3.27468779155936,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.7562522888183594,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.02466920834267773,
546
+ "grad_norm": 7.595749246101231,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.7546162605285645,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.024989587671803416,
553
+ "grad_norm": 2.1701956042711728,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.7503818511962891,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.0253099670009291,
560
+ "grad_norm": 7.551313750149006,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.7194967269897461,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.025630346330054784,
567
+ "grad_norm": 3.0982279258014085,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.7183728218078613,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.025950725659180468,
574
+ "grad_norm": 2.2319456779746507,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.741051197052002,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.026271104988306156,
581
+ "grad_norm": 2.1272967612583904,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.7074191093444824,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.02659148431743184,
588
+ "grad_norm": 2.3756745796722125,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.7466435909271241,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.026911863646557524,
595
+ "grad_norm": 3.9008908753860925,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.7262076377868653,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.027232242975683208,
602
+ "grad_norm": 2.0561831766929806,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.7282081604003906,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.027552622304808892,
609
+ "grad_norm": 5.818198655017998,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.7566670894622802,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.02787300163393458,
616
+ "grad_norm": 2.1494490230827337,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.7288908958435059,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.028193380963060264,
623
+ "grad_norm": 1.409242496714709,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.7215320587158203,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.028513760292185948,
630
+ "grad_norm": 7.0920590147047164,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.7246700286865234,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.028834139621311632,
637
+ "grad_norm": 1.5684095793164903,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.7144109725952148,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.029154518950437316,
644
+ "grad_norm": 1.3316395400018082,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.7477982997894287,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.029474898279563004,
651
+ "grad_norm": 1.3602729023468434,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.7173037528991699,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.029795277608688688,
658
+ "grad_norm": 49.53643391306756,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.7678704738616944,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.030115656937814372,
665
+ "grad_norm": 4.340394359223178,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.7031218528747558,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.030436036266940057,
672
+ "grad_norm": 2.1203555620583736,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.7084536552429199,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.03075641559606574,
679
+ "grad_norm": 1.3371075573755178,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.7464202880859375,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.03107679492519143,
686
+ "grad_norm": 1.6473009838115498,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.7116725444793701,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.03139717425431711,
693
+ "grad_norm": 3.385121877804678,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.738076114654541,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.0317175535834428,
700
+ "grad_norm": 1.7147230699245448,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.7158563137054443,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.032037932912568484,
707
+ "grad_norm": 3.3445220753393468,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.8030566215515137,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.032358312241694165,
714
+ "grad_norm": 2.077502109986157,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.7472703456878662,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.03267869157081985,
721
+ "grad_norm": 17.329143506946714,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.7366230964660645,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.03299907089994553,
728
+ "grad_norm": 2.6268238025254735,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.7404189109802246,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.03331945022907122,
735
+ "grad_norm": 1.977866371730136,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.7236545085906982,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.03363982955819691,
742
+ "grad_norm": 1.498659413315871,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.7335420608520508,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.03396020888732259,
749
+ "grad_norm": 1.7062214656423396,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.7051788330078125,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.03428058821644828,
756
+ "grad_norm": 1.3397922877558162,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.7626142024993896,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.03460096754557396,
763
+ "grad_norm": 3.93050291862445,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.7514806270599366,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.034921346874699645,
770
+ "grad_norm": 3.398990137450838,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.7132882595062255,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.03524172620382533,
777
+ "grad_norm": 1.4600798546048035,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.716385793685913,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.03556210553295101,
784
+ "grad_norm": 1.3160301135986112,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.7191272735595703,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.0358824848620767,
791
+ "grad_norm": 1.2481731524679547,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.7158429145812988,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.03620286419120238,
798
+ "grad_norm": 2.8179317711592504,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.7038905143737793,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.03652324352032807,
805
+ "grad_norm": 1.8284267025580432,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.7065417289733886,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.03684362284945375,
812
+ "grad_norm": 2.212892645914091,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.7265023708343505,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.03716400217857944,
819
+ "grad_norm": 2.2597923858522404,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.7199366569519043,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.037484381507705125,
826
+ "grad_norm": 2.0160695548330936,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.7658578395843506,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.037804760836830806,
833
+ "grad_norm": 2.5572488896871843,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.7220034599304199,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.03812514016595649,
840
+ "grad_norm": 2.459122478774182,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.6983201026916503,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.038445519495082174,
847
+ "grad_norm": 5.318818273924634,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.7152614116668701,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.03876589882420786,
854
+ "grad_norm": 4.348661337562832,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.7531236171722412,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.03908627815333355,
861
+ "grad_norm": 3.5073096880649666,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.7070003509521484,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.03940665748245923,
868
+ "grad_norm": 6.250765626260255,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.7422648429870605,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.03972703681158492,
875
+ "grad_norm": 2.1134272476858693,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.7167387962341308,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.0400474161407106,
882
+ "grad_norm": 2.728613864385018,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.7613626480102539,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.040367795469836286,
889
+ "grad_norm": 2.8289243765180094,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.7406154155731202,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.040688174798961974,
896
+ "grad_norm": 1.460978022101602,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.7171460628509522,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.041008554128087654,
903
+ "grad_norm": 11.187948658326427,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.7119218349456787,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.04132893345721334,
910
+ "grad_norm": 1.5366780655953838,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.7250003337860107,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.04164931278633902,
917
+ "grad_norm": 2.6451401695263392,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.7535340785980225,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.04196969211546471,
924
+ "grad_norm": 5.247231775991024,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.7120182991027832,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.0422900714445904,
931
+ "grad_norm": 1.265414487137248,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.7114714622497559,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.04261045077371608,
938
+ "grad_norm": 32.52328999465322,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.6845267295837403,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.042930830102841766,
945
+ "grad_norm": 6.1768247607509394,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.702857780456543,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.04325120943196745,
952
+ "grad_norm": 7.1933410968196725,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.72010817527771,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.043571588761093134,
959
+ "grad_norm": 1.1755562476246728,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.6978877067565918,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.04389196809021882,
966
+ "grad_norm": 2.8540830213502844,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.7204657554626465,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.0442123474193445,
973
+ "grad_norm": 1.1871852286932287,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.7182430267333985,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.04453272674847019,
980
+ "grad_norm": 1.0566209393736636,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.7123696804046631,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.04485310607759587,
987
+ "grad_norm": 4.237024261846737,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.6963154792785644,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.04517348540672156,
994
+ "grad_norm": 2.008976090493564,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.6849660396575927,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.045493864735847246,
1001
+ "grad_norm": 1.3493113476458192,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.7087863922119141,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.04581424406497293,
1008
+ "grad_norm": 1.0453265260241618,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.6820076942443848,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.046134623394098614,
1015
+ "grad_norm": 1.367769743280532,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.7149313926696778,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.046455002723224295,
1022
+ "grad_norm": 1.5424097504581287,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.7253753185272217,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.04677538205234998,
1029
+ "grad_norm": 2.3372729347553025,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.7671535015106201,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.04709576138147567,
1036
+ "grad_norm": 1.8742828809078558,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.7182416915893555,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.04741614071060135,
1043
+ "grad_norm": 6.87172572472952,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.7154425621032715,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.04773652003972704,
1050
+ "grad_norm": 2.4729840403642545,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.7049578666687012,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.04805689936885272,
1057
+ "grad_norm": 1.3804743410158802,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.7238261222839355,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.04837727869797841,
1064
+ "grad_norm": 1.2556616325656202,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.7582721710205078,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.048697658027104095,
1071
+ "grad_norm": 1.4068662819395166,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.6943454742431641,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.049018037356229775,
1078
+ "grad_norm": 1.5191574570193485,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.7089879512786865,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.04933841668535546,
1085
+ "grad_norm": 1.1100647663489502,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.6814855098724365,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.049658796014481144,
1092
+ "grad_norm": 1.2368089337986163,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.7276053428649902,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.04997917534360683,
1099
+ "grad_norm": 1.596887157213148,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.7387615203857422,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.05029955467273251,
1106
+ "grad_norm": 1.3857783026385584,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.6814603805541992,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.0506199340018582,
1113
+ "grad_norm": 1.5717792680161595,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.6634132385253906,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.05094031333098389,
1120
+ "grad_norm": 5.076622025769138,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.7170121669769287,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.05126069266010957,
1127
+ "grad_norm": 3.075120261257635,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.6901824951171875,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.051581071989235255,
1134
+ "grad_norm": 3.1443073556392838,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.7126097679138184,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.051901451318360936,
1141
+ "grad_norm": 1.3279065854265553,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.696572208404541,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.052221830647486624,
1148
+ "grad_norm": 2.2922495579972213,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.7236587047576905,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.05254220997661231,
1155
+ "grad_norm": 1.853599288786833,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.6959040641784668,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.05286258930573799,
1162
+ "grad_norm": 1.6356938422959257,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.6883045196533203,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.05318296863486368,
1169
+ "grad_norm": 1.4258158618988481,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.7095949172973632,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.05350334796398936,
1176
+ "grad_norm": 1.3818626048983265,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.7279641628265381,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.05382372729311505,
1183
+ "grad_norm": 1.3148725391903133,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.7071706295013428,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.054144106622240736,
1190
+ "grad_norm": 1.3205843385485543,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.7102875709533691,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.054464485951366416,
1197
+ "grad_norm": 1.0245118537391376,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.7355536460876465,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.054784865280492104,
1204
+ "grad_norm": 0.9830319826014532,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.711072301864624,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.055105244609617784,
1211
+ "grad_norm": 2.2427073223135,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.7257272720336914,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.05542562393874347,
1218
+ "grad_norm": 1.7432652030049554,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.7266391754150391,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.05574600326786916,
1225
+ "grad_norm": 7.3855085380936965,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.7229394435882568,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.05606638259699484,
1232
+ "grad_norm": 2.1417039515536542,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.6833599090576172,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.05638676192612053,
1239
+ "grad_norm": 2.4468519837946165,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.7663132667541503,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.05670714125524621,
1246
+ "grad_norm": 3.0137174378050293,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.7224421977996827,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.057027520584371896,
1253
+ "grad_norm": 2.2523438896707706,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.7207423210144043,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.057347899913497584,
1260
+ "grad_norm": 2.919705865322085,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.6854898452758789,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.057668279242623265,
1267
+ "grad_norm": 3.984396788478948,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.7234033107757568,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.05798865857174895,
1274
+ "grad_norm": 6.67851094941151,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.7022351264953614,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.05830903790087463,
1281
+ "grad_norm": 1.6866126836675677,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.7415154457092286,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.05862941723000032,
1288
+ "grad_norm": 3.218820869797509,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.7005694389343262,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.05894979655912601,
1295
+ "grad_norm": 3.0737908739781106,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.6981851100921631,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.05927017588825169,
1302
+ "grad_norm": 2.2327032848527795,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.7604454040527344,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.059590555217377376,
1309
+ "grad_norm": 2.8761311961580973,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.6888795852661133,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.05991093454650306,
1316
+ "grad_norm": 2.2317201691382533,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.708838415145874,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.060231313875628745,
1323
+ "grad_norm": 2.0355186191234806,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.7040927886962891,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.06055169320475443,
1330
+ "grad_norm": 2.014047361446366,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.7197917461395263,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.06087207253388011,
1337
+ "grad_norm": 1.977104681851146,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.6776611328125,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.0611924518630058,
1344
+ "grad_norm": 13.901447338457938,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.6899932861328125,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.06151283119213148,
1351
+ "grad_norm": 2.1123993495266435,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.6683700084686279,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.06183321052125717,
1358
+ "grad_norm": 2.3306798040564676,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.691620922088623,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.06215358985038286,
1365
+ "grad_norm": 13.395393364218384,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.7234976768493653,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.06247396917950854,
1372
+ "grad_norm": 7.087449907589583,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.7050728797912598,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.06279434850863422,
1379
+ "grad_norm": 2.912496650205425,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.6966752529144287,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.06311472783775991,
1386
+ "grad_norm": 8.623415554665941,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.67595534324646,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.0634351071668856,
1393
+ "grad_norm": 3.239938193275378,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.6972695350646972,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.06375548649601127,
1400
+ "grad_norm": 4.165992500546785,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.7464128017425538,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.06407586582513697,
1407
+ "grad_norm": 3.308146075178026,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.7509006023406982,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.06439624515426265,
1414
+ "grad_norm": 3.4969526861271802,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.7300586700439453,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.06471662448338833,
1421
+ "grad_norm": 4.743378784447764,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.6886860847473144,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.06503700381251401,
1428
+ "grad_norm": 1.9438372463888733,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.705808448791504,
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:e12d9e9836f90b7e3bf6e6beadff81e2cd76af2fbd6aa526cb1396f85ef63153
3
+ size 7313