Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-g2_2_oldcode_synthcm_newcode

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
3b091d4
·
verified ·
1 Parent(s): 6ed2a9b

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 sha256:927aad38f52ee9a73094c34471ceb31f4779ca060be109acb2cb6e289b81ab21
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": 0.042687531465010906,
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": 0.00020976673938580297,
14
+ "grad_norm": 236.7567785411443,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 1.5920657157897948,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.00041953347877160595,
21
+ "grad_norm": 152.92314363522394,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 0.9421598434448242,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.000629300218157409,
28
+ "grad_norm": 97.19978632147773,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.7996736526489258,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.0008390669575432119,
35
+ "grad_norm": 157.7423675487546,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.760670280456543,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.001048833696929015,
42
+ "grad_norm": 49.50204924369943,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.7635975837707519,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.001258600436314818,
49
+ "grad_norm": 16.48624285559217,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.7791356563568115,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0014683671757006209,
56
+ "grad_norm": 6.858907596379825,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.7495433330535889,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.0016781339150864238,
63
+ "grad_norm": 26.435928023185784,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.7318498134613037,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.001887900654472227,
70
+ "grad_norm": 214.57004030820568,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.7413080215454102,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.00209766739385803,
77
+ "grad_norm": 7.939148544393301,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.8061410903930664,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.0023074341332438328,
84
+ "grad_norm": 14.183647299272815,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.7331802368164062,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.002517200872629636,
91
+ "grad_norm": 22.423640491145502,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.7498276710510254,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.002726967612015439,
98
+ "grad_norm": 3.136265232048693,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.724825668334961,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.0029367343514012417,
105
+ "grad_norm": 8.214245546263463,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 0.6886782646179199,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.003146501090787045,
112
+ "grad_norm": 5.619530923664304,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.7081296443939209,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.0033562678301728476,
119
+ "grad_norm": 10.033723007819054,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.7299324035644531,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.0035660345695586507,
126
+ "grad_norm": 11.134808180954863,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.745941162109375,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.003775801308944454,
133
+ "grad_norm": 2.4304907872424795,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.7129830360412598,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.0039855680483302566,
140
+ "grad_norm": 4.018584134102583,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.7092458724975585,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.00419533478771606,
147
+ "grad_norm": 5.006464475899677,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.7565587520599365,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.004405101527101863,
154
+ "grad_norm": 2.800599112711121,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.6959824562072754,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.0046148682664876655,
161
+ "grad_norm": 6.231621527897506,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.6992167472839356,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.004824635005873469,
168
+ "grad_norm": 10.811727778354081,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.6952213287353516,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.005034401745259272,
175
+ "grad_norm": 11.261067852316236,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.7460278987884521,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.0052441684846450745,
182
+ "grad_norm": 7.649125680617738,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.7858521461486816,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.005453935224030878,
189
+ "grad_norm": 4.490247060675485,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.7642570018768311,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.005663701963416681,
196
+ "grad_norm": 5.320163840820431,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.7289691925048828,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.0058734687028024835,
203
+ "grad_norm": 3.4799647445683717,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.6995810508728028,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.006083235442188286,
210
+ "grad_norm": 34.442893954128465,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.7102492809295654,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.00629300218157409,
217
+ "grad_norm": 4.179341348142214,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.7110370635986328,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.0065027689209598925,
224
+ "grad_norm": 4.270936726323635,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.7388000011444091,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.006712535660345695,
231
+ "grad_norm": 3.52144124715383,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.7715152740478516,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.006922302399731499,
238
+ "grad_norm": 2.1760255438109892,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.7474425315856934,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.0071320691391173014,
245
+ "grad_norm": 1.8391052662069023,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.7086240291595459,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.007341835878503104,
252
+ "grad_norm": 9.85499125185226,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.736334228515625,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.007551602617888908,
259
+ "grad_norm": 7.061981460856605,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.7058637619018555,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.00776136935727471,
266
+ "grad_norm": 2.399773463771947,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.7035298824310303,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.007971136096660513,
273
+ "grad_norm": 1.8715337008198967,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.7319884300231934,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.008180902836046316,
280
+ "grad_norm": 4.402380567892121,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.6931228637695312,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.00839066957543212,
287
+ "grad_norm": 1.997385308499001,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.6841664791107178,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.008600436314817923,
294
+ "grad_norm": 8.77520106541026,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.6674545288085938,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.008810203054203726,
301
+ "grad_norm": 227.00469733968774,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.7310820579528808,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.009019969793589528,
308
+ "grad_norm": 3.283083883742064,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.7967383384704589,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.009229736532975331,
315
+ "grad_norm": 12.336376030486466,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.7576329231262207,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.009439503272361134,
322
+ "grad_norm": 1.8242777860324726,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.706461238861084,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.009649270011746938,
329
+ "grad_norm": 1.8090965567853239,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.7032674312591553,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.009859036751132741,
336
+ "grad_norm": 2.5475991670097953,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.7029439449310303,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.010068803490518544,
343
+ "grad_norm": 1.4692934900457209,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.6630709648132325,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.010278570229904346,
350
+ "grad_norm": 2.1331924790853534,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.7047128677368164,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.010488336969290149,
357
+ "grad_norm": 2.4224465959516164,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.7062530517578125,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.010698103708675952,
364
+ "grad_norm": 1.9672184607046945,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.6654829978942871,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.010907870448061756,
371
+ "grad_norm": 1.8632555104139144,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.6872918128967285,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.011117637187447559,
378
+ "grad_norm": 2.696765708839419,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.712721586227417,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.011327403926833362,
385
+ "grad_norm": 3.3114397630142145,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.665414571762085,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.011537170666219164,
392
+ "grad_norm": 1.7287592058057928,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.7328654289245605,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.011746937405604967,
399
+ "grad_norm": 1.7622716775547056,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.692367696762085,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.01195670414499077,
406
+ "grad_norm": 2.3708233226098807,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.6588814735412598,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.012166470884376572,
413
+ "grad_norm": 4.863667938523831,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.7014576911926269,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.012376237623762377,
420
+ "grad_norm": 2.1632337576331073,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.7060187816619873,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.01258600436314818,
427
+ "grad_norm": 2.1446479504758584,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.7127746105194092,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.012795771102533982,
434
+ "grad_norm": 2.103648073321234,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.7010409355163574,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.013005537841919785,
441
+ "grad_norm": 2.300281953023738,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.7145485877990723,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.013215304581305588,
448
+ "grad_norm": 2.1745603402017855,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.6661243438720703,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.01342507132069139,
455
+ "grad_norm": 3.3911386633188085,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.7290327072143554,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.013634838060077195,
462
+ "grad_norm": 2.838234667045874,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.7416918754577637,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.013844604799462997,
469
+ "grad_norm": 2.6030763255310494,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.6962085247039795,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.0140543715388488,
476
+ "grad_norm": 3.9879154447494263,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.7320832729339599,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.014264138278234603,
483
+ "grad_norm": 5.147234411879652,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.690255880355835,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.014473905017620406,
490
+ "grad_norm": 4.41792994494731,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.7107495784759521,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.014683671757006208,
497
+ "grad_norm": 4.0092748239949465,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.7144205093383789,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.014893438496392013,
504
+ "grad_norm": 3.663722199190995,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.678815221786499,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.015103205235777815,
511
+ "grad_norm": 374.6324936776676,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.7061020851135253,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.015312971975163618,
518
+ "grad_norm": 2.830711027072347,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.716461181640625,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.01552273871454942,
525
+ "grad_norm": 17.687806831387658,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.6958938598632812,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.015732505453935224,
532
+ "grad_norm": 67.22665881659304,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.6571521759033203,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.015942272193321026,
539
+ "grad_norm": 3.2787186518009888,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.737470006942749,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.01615203893270683,
546
+ "grad_norm": 5.764665619355419,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.6784905433654785,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.01636180567209263,
553
+ "grad_norm": 2.8187442313274156,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.698319387435913,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.016571572411478434,
560
+ "grad_norm": 11.052681670850987,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.7136563301086426,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.01678133915086424,
567
+ "grad_norm": 2.738088534726021,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.7217898368835449,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.016991105890250043,
574
+ "grad_norm": 22.99130603074463,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.6949929237365723,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.017200872629635846,
581
+ "grad_norm": 2.2673846343594293,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.6527021408081055,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.01741063936902165,
588
+ "grad_norm": 4.040669613703935,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.6478307723999024,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.01762040610840745,
595
+ "grad_norm": 1.7450269280537714,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.7058829307556153,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.017830172847793254,
602
+ "grad_norm": 1.3888768876108615,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.6917304039001465,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.018039939587179057,
609
+ "grad_norm": 2.497192781134032,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.6497130393981934,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.01824970632656486,
616
+ "grad_norm": 2.096909255558357,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.6941553115844726,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.018459473065950662,
623
+ "grad_norm": 1.7631454307114556,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.7313367366790772,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.018669239805336465,
630
+ "grad_norm": 1.2137215799580678,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.734074592590332,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.018879006544722268,
637
+ "grad_norm": 8.527470156498259,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.7139863967895508,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.01908877328410807,
644
+ "grad_norm": 1.150419868331103,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.7071406364440918,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.019298540023493876,
651
+ "grad_norm": 2.4598251273735596,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.7090801239013672,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.01950830676287968,
658
+ "grad_norm": 2.0085089193495707,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.6835993766784668,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.019718073502265482,
665
+ "grad_norm": 3.0778601053919536,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.7152371883392334,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.019927840241651285,
672
+ "grad_norm": 6.504260026804486,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.6743127346038819,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.020137606981037087,
679
+ "grad_norm": 1.7969857643634526,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.6778482437133789,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.02034737372042289,
686
+ "grad_norm": 29.433979140864412,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.6873385429382324,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.020557140459808693,
693
+ "grad_norm": 2.035543156408046,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.6699578762054443,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.020766907199194495,
700
+ "grad_norm": 1.1357536583710182,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.695745849609375,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.020976673938580298,
707
+ "grad_norm": 3.7046794359871447,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.6789576530456543,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.0211864406779661,
714
+ "grad_norm": 3.130957938855976,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.6930006980895996,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.021396207417351903,
721
+ "grad_norm": 11.276914781871405,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.6952384948730469,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.021605974156737706,
728
+ "grad_norm": 6.337861678743805,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.6785932064056397,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.021815740896123512,
735
+ "grad_norm": 4.784777304261159,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.7035085678100585,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.022025507635509315,
742
+ "grad_norm": 5.679658714316126,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.6616763114929199,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.022235274374895118,
749
+ "grad_norm": 3.478649303767567,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.6685798645019532,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.02244504111428092,
756
+ "grad_norm": 5.634898233949741,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.7499291896820068,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.022654807853666723,
763
+ "grad_norm": 5.773543600338411,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.6558557033538819,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.022864574593052526,
770
+ "grad_norm": 39.700232901357666,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.6942555904388428,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.02307434133243833,
777
+ "grad_norm": 6.2820558818505265,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.6771759033203125,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.02328410807182413,
784
+ "grad_norm": 2.9316401080763823,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.6969615936279296,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.023493874811209934,
791
+ "grad_norm": 2.8974904226392466,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.6931879997253418,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.023703641550595737,
798
+ "grad_norm": 4.189761025368283,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.6718255043029785,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.02391340828998154,
805
+ "grad_norm": 2.417352201543177,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.7471983909606934,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.024123175029367342,
812
+ "grad_norm": 2.839365825197699,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.6793410301208496,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.024332941768753145,
819
+ "grad_norm": 3.141367299942584,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.7086249351501465,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.02454270850813895,
826
+ "grad_norm": 2.060642183486379,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.6504176139831543,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.024752475247524754,
833
+ "grad_norm": 23.95719687060173,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.6888862609863281,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.024962241986910556,
840
+ "grad_norm": 5.209642741753617,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.6663257598876953,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.02517200872629636,
847
+ "grad_norm": 4.668122000612013,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.6789710521697998,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.025381775465682162,
854
+ "grad_norm": 2.21317699920472,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.6676199913024903,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.025591542205067964,
861
+ "grad_norm": 4.368293636030707,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.690043592453003,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.025801308944453767,
868
+ "grad_norm": 3.8199119576336176,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.7009891033172607,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.02601107568383957,
875
+ "grad_norm": 2.3977516666695142,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.6778505325317383,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.026220842423225373,
882
+ "grad_norm": 12.32747378491102,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.7028230190277099,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.026430609162611175,
889
+ "grad_norm": 5.6463852018519844,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.6816394329071045,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.026640375901996978,
896
+ "grad_norm": 3.0036641576788443,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.6653374671936035,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.02685014264138278,
903
+ "grad_norm": 3.963136105803816,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.70212721824646,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.027059909380768587,
910
+ "grad_norm": 2.3922579167108338,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.6652139186859131,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.02726967612015439,
917
+ "grad_norm": 1.430307897978201,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.649704122543335,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.027479442859540192,
924
+ "grad_norm": 12.579409728745457,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.7093433380126953,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.027689209598925995,
931
+ "grad_norm": 1.8768014730990727,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.6947019100189209,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.027898976338311798,
938
+ "grad_norm": 4.087020619174566,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.6580727577209473,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.0281087430776976,
945
+ "grad_norm": 2.2104247401520505,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.6761454582214356,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.028318509817083403,
952
+ "grad_norm": 1.54359381136085,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.7064968109130859,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.028528276556469206,
959
+ "grad_norm": 9.445161276396444,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.669188404083252,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.02873804329585501,
966
+ "grad_norm": 2.7373325593304836,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.6705168724060059,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.02894781003524081,
973
+ "grad_norm": 2.276476538972258,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.7053794860839844,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.029157576774626614,
980
+ "grad_norm": 9.277889149036925,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.6553175926208497,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.029367343514012417,
987
+ "grad_norm": 3.368267525079364,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.6869246959686279,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.029577110253398223,
994
+ "grad_norm": 3.2226378523733645,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.6814833641052246,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.029786876992784025,
1001
+ "grad_norm": 1.930571790233891,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.6552035331726074,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.029996643732169828,
1008
+ "grad_norm": 3.5201808130997003,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.6717945098876953,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.03020641047155563,
1015
+ "grad_norm": 660.3628631506493,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.6539801597595215,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.030416177210941434,
1022
+ "grad_norm": 5.251919957664849,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.6830947875976563,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.030625943950327236,
1029
+ "grad_norm": 5.249421083313879,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.681186580657959,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.03083571068971304,
1036
+ "grad_norm": 2.0509657202306752,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.6720983028411865,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.03104547742909884,
1043
+ "grad_norm": 2.2734683613081064,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.67911696434021,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.031255244168484644,
1050
+ "grad_norm": 1.4795523059161764,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.6568414211273194,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.03146501090787045,
1057
+ "grad_norm": 1.2453772205405704,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.6762202262878418,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.03167477764725625,
1064
+ "grad_norm": 2.0010667543841123,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.6813722133636475,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.03188454438664205,
1071
+ "grad_norm": 39.56264214911426,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.7298669815063477,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.032094311126027855,
1078
+ "grad_norm": 1.9507720129351338,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.6247622489929199,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.03230407786541366,
1085
+ "grad_norm": 1.484907363192865,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.6759968757629394,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.03251384460479946,
1092
+ "grad_norm": 1.9147431404888071,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.6800020217895508,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.03272361134418526,
1099
+ "grad_norm": 3.5428313404092,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.6901631355285645,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.032933378083571066,
1106
+ "grad_norm": 4.7250773473958105,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.6709762573242187,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.03314314482295687,
1113
+ "grad_norm": 5.898810367933252,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.66232271194458,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.03335291156234268,
1120
+ "grad_norm": 1.9369336440014253,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.7334044933319092,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.03356267830172848,
1127
+ "grad_norm": 8.737915524161552,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.6792474269866944,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.033772445041114284,
1134
+ "grad_norm": 1.9836799556383249,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.7033929824829102,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.033982211780500086,
1141
+ "grad_norm": 1.1854978830042706,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.6635163307189942,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.03419197851988589,
1148
+ "grad_norm": 1.2197328294566572,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.6634726524353027,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.03440174525927169,
1155
+ "grad_norm": 1.4709834426885486,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.6794222831726074,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.034611511998657495,
1162
+ "grad_norm": 1.5366826867181045,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.6605133056640625,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.0348212787380433,
1169
+ "grad_norm": 2.9147800494231335,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.6418501853942871,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.0350310454774291,
1176
+ "grad_norm": 1.202686476162776,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.633223295211792,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.0352408122168149,
1183
+ "grad_norm": 0.9631875057504979,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.666954517364502,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.035450578956200705,
1190
+ "grad_norm": 1.4192365422706907,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.6560389518737793,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.03566034569558651,
1197
+ "grad_norm": 1.1713801817552898,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.641888427734375,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.03587011243497231,
1204
+ "grad_norm": 1.4022563568499706,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.7102468013763428,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.036079879174358113,
1211
+ "grad_norm": 2.0986091486389826,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.6136519908905029,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.036289645913743916,
1218
+ "grad_norm": 1.1634525834102774,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.6924193382263184,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.03649941265312972,
1225
+ "grad_norm": 2.081748039936143,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.6236453533172608,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.03670917939251552,
1232
+ "grad_norm": 2.484079612019191,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.6446088314056396,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.036918946131901324,
1239
+ "grad_norm": 2.142979943466157,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.6566426753997803,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.03712871287128713,
1246
+ "grad_norm": 2.727971395956995,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.6372243404388428,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.03733847961067293,
1253
+ "grad_norm": 74.34931845644519,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.6198902130126953,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.03754824635005873,
1260
+ "grad_norm": 2.066092794611252,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.6555921554565429,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.037758013089444535,
1267
+ "grad_norm": 3.0945696621385244,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.6903239250183105,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.03796777982883034,
1274
+ "grad_norm": 2.0553070965106786,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.6789304733276367,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.03817754656821614,
1281
+ "grad_norm": 1.9236355838089434,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.6835772514343261,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.03838731330760194,
1288
+ "grad_norm": 1.1647199736335656,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.6774552822113037,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.03859708004698775,
1295
+ "grad_norm": 5.471968133112629,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.6373446464538575,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.038806846786373556,
1302
+ "grad_norm": 1.4710591021325636,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.6392849922180176,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.03901661352575936,
1309
+ "grad_norm": 2.6999041311336756,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.6585689067840577,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.03922638026514516,
1316
+ "grad_norm": 1.1333104342309857,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.6750880241394043,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.039436147004530964,
1323
+ "grad_norm": 1.8072514966793143,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.6438388824462891,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.039645913743916766,
1330
+ "grad_norm": 1.0781542511331554,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.6553819179534912,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.03985568048330257,
1337
+ "grad_norm": 3.0719828705316616,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.6419807434082031,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.04006544722268837,
1344
+ "grad_norm": 1.5436345848872006,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.6960256099700928,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.040275213962074174,
1351
+ "grad_norm": 1.0011408997937439,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.6509086608886718,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.04048498070145998,
1358
+ "grad_norm": 1.294803313028635,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.6446637153625489,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.04069474744084578,
1365
+ "grad_norm": 1.4561867936361599,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.6471066474914551,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.04090451418023158,
1372
+ "grad_norm": 1.420758879294619,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.6668454647064209,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.041114280919617385,
1379
+ "grad_norm": 1.339731233422897,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.6452267646789551,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.04132404765900319,
1386
+ "grad_norm": 3.1201962896167976,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.643893814086914,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.04153381439838899,
1393
+ "grad_norm": 1.724067302763631,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.6555222988128662,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.04174358113777479,
1400
+ "grad_norm": 2.166362359535913,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.6665771484375,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.041953347877160596,
1407
+ "grad_norm": 2.744806451285816,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.6609567165374756,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.0421631146165464,
1414
+ "grad_norm": 1.1718498443702758,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.6552879333496093,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.0423728813559322,
1421
+ "grad_norm": 20.838352988266458,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.6685161590576172,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.042582648095318004,
1428
+ "grad_norm": 1.0046844547391474,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.6807273864746094,
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:f5c99a7546b881ae02687e4a30d32fb2e5e1a1b2df0c529f0558c661b7a3fac2
3
+ size 7313