Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-ov_C_repeat

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
a641ed3
·
verified ·
1 Parent(s): 9682c6c

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 sha256:2ac38e26ba82179b05c78220dae64351ecbc101c624a301cfc0a3d7c64b21b55
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": 0.26150089951169364,
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": 0.0012850167052171678,
14
+ "grad_norm": 856.6189574701012,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 1.1357397079467773,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0025700334104343356,
21
+ "grad_norm": 15.552391128873273,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 0.7902592658996582,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.0038550501156515036,
28
+ "grad_norm": 15.294031112683584,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.7169486045837402,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.005140066820868671,
35
+ "grad_norm": 14.61793688455095,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.6879290580749512,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.006425083526085839,
42
+ "grad_norm": 17.865572565928804,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.6837553977966309,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.007710100231303007,
49
+ "grad_norm": 78.02976802685448,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.6830409049987793,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.008995116936520175,
56
+ "grad_norm": 13.787148293237584,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.6678519725799561,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.010280133641737342,
63
+ "grad_norm": 59.88701484053759,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.671458625793457,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.01156515034695451,
70
+ "grad_norm": 14.98452507287525,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.6946072578430176,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.012850167052171678,
77
+ "grad_norm": 9.949341410744578,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.6994287014007569,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.014135183757388847,
84
+ "grad_norm": 20.083005842913845,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.654233169555664,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.015420200462606014,
91
+ "grad_norm": 52.69058821223778,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.6701527595520019,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.01670521716782318,
98
+ "grad_norm": 33.06654632843674,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.6579795837402344,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.01799023387304035,
105
+ "grad_norm": 26.181216707950213,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 0.6791215419769288,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.01927525057825752,
112
+ "grad_norm": 25.305185028986585,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.6644459724426269,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.020560267283474685,
119
+ "grad_norm": 4.88393687591383,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.6924647808074951,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.021845283988691854,
126
+ "grad_norm": 7.644492046355705,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.6735517024993897,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.02313030069390902,
133
+ "grad_norm": 4.0610145968849345,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.6673793315887451,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.02441531739912619,
140
+ "grad_norm": 9.704924216779212,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.6797656059265137,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.025700334104343355,
147
+ "grad_norm": 5.615408309945265,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.684023666381836,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.026985350809560524,
154
+ "grad_norm": 5.423083132935791,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.6746089935302735,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.028270367514777694,
161
+ "grad_norm": 8.79838661662345,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.6778485774993896,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.02955538421999486,
168
+ "grad_norm": 3.389056327858364,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.6816946029663086,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.03084040092521203,
175
+ "grad_norm": 3.2806365569635534,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.6739953517913818,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.0321254176304292,
182
+ "grad_norm": 2.714387951683758,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.6525035858154297,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.03341043433564636,
189
+ "grad_norm": 3.158516093565955,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.6718252182006836,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.03469545104086353,
196
+ "grad_norm": 5.772344202126518,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.647340965270996,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.0359804677460807,
203
+ "grad_norm": 5.502021173920203,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.6496793746948242,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.03726548445129787,
210
+ "grad_norm": 2.68008821860417,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.6709864616394043,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.03855050115651504,
217
+ "grad_norm": 1.7973533832528188,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.6566804885864258,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.0398355178617322,
224
+ "grad_norm": 6.51867504563315,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.6657544136047363,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.04112053456694937,
231
+ "grad_norm": 2.362613498811483,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.6502259731292724,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.04240555127216654,
238
+ "grad_norm": 3.907152138683563,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.6525882244110107,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.04369056797738371,
245
+ "grad_norm": 3.794880030440209,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.6532615661621094,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.04497558468260087,
252
+ "grad_norm": 8.769127576551398,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.6508918285369873,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.04626060138781804,
259
+ "grad_norm": 2.239128040127585,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.6526441097259521,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.04754561809303521,
266
+ "grad_norm": 3.6055555343647714,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.681905460357666,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.04883063479825238,
273
+ "grad_norm": 2.4554667315087624,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.6597890377044677,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.05011565150346955,
280
+ "grad_norm": 1.8867583686163962,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.6532648086547852,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.05140066820868671,
287
+ "grad_norm": 1.3820672524188946,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.653763723373413,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.05268568491390388,
294
+ "grad_norm": 4.353588095279891,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.6449775695800781,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.05397070161912105,
301
+ "grad_norm": 4.764374381297121,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.6555165767669677,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.05525571832433822,
308
+ "grad_norm": 14.727196337345648,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.6529878139495849,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.05654073502955539,
315
+ "grad_norm": 3.744804024990321,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.6880834579467774,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.05782575173477255,
322
+ "grad_norm": 9.9949909991041,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.6793588161468506,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.05911076843998972,
329
+ "grad_norm": 4.6856351506601595,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.6606217384338379,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.06039578514520689,
336
+ "grad_norm": 2.5156498724259575,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.6711392879486084,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.06168080185042406,
343
+ "grad_norm": 3.205814816900315,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.6852891921997071,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.06296581855564122,
350
+ "grad_norm": 1.839545887284569,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.6825028896331787,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.0642508352608584,
357
+ "grad_norm": 2.4792665803649783,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.6519015789031982,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.06553585196607556,
364
+ "grad_norm": 1.9240510741369987,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.648968505859375,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.06682086867129272,
371
+ "grad_norm": 2.5695541275375082,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.6466852188110351,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.0681058853765099,
378
+ "grad_norm": 11.336408841436898,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.6370265483856201,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.06939090208172706,
385
+ "grad_norm": 2.7714988957574405,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.6360719680786133,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.07067591878694424,
392
+ "grad_norm": 2.5399185097295414,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.6560811996459961,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.0719609354921614,
399
+ "grad_norm": 4.439208424776135,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.644292163848877,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.07324595219737856,
406
+ "grad_norm": 5.696317025602095,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.6401437282562256,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.07453096890259574,
413
+ "grad_norm": 2.664013724679843,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.6491299629211426,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.0758159856078129,
420
+ "grad_norm": 2.1745058771672237,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.63185133934021,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.07710100231303008,
427
+ "grad_norm": 3.0153668076131943,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.6398284912109375,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.07838601901824724,
434
+ "grad_norm": 4.211123263285935,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.645693016052246,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.0796710357234644,
441
+ "grad_norm": 9.078438471416925,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.6547775268554688,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.08095605242868158,
448
+ "grad_norm": 1.5458422123622622,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.6409534454345703,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.08224106913389874,
455
+ "grad_norm": 6.94798620371141,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.6655917644500733,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.0835260858391159,
462
+ "grad_norm": 1.271505509202168,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.6486823558807373,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.08481110254433308,
469
+ "grad_norm": 1.0317616781253849,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.6482730388641358,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.08609611924955024,
476
+ "grad_norm": 1.1601265249889448,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.6461831569671631,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.08738113595476742,
483
+ "grad_norm": 1.5955812719491171,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.6477439403533936,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.08866615265998458,
490
+ "grad_norm": 1.2974816270859992,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.630802059173584,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.08995116936520174,
497
+ "grad_norm": 1.802080446959691,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.6312804698944092,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.09123618607041892,
504
+ "grad_norm": 1.2734807542979136,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.6359667778015137,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.09252120277563608,
511
+ "grad_norm": 1.089715475560824,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.6190359115600585,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.09380621948085326,
518
+ "grad_norm": 1.4111528747217608,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.6462649822235107,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.09509123618607042,
525
+ "grad_norm": 2.3575893938615566,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.6176736354827881,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.09637625289128758,
532
+ "grad_norm": 4.099410091763762,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.6239109039306641,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.09766126959650476,
539
+ "grad_norm": 4.853333713828358,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.6444282531738281,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.09894628630172192,
546
+ "grad_norm": 2.5867597289295405,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.6411311626434326,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.1002313030069391,
553
+ "grad_norm": 3.048638070554145,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.6209658622741699,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.10151631971215626,
560
+ "grad_norm": 1.9668189975150188,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.6313361167907715,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.10280133641737342,
567
+ "grad_norm": 2.7558941074380496,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.6456878185272217,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.1040863531225906,
574
+ "grad_norm": 4.931251393074305,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.616514778137207,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.10537136982780776,
581
+ "grad_norm": 1.9224203127091826,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.6402089595794678,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.10665638653302494,
588
+ "grad_norm": 2.6967273056980985,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.6169169425964356,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.1079414032382421,
595
+ "grad_norm": 2.935113359723491,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.6237776756286622,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.10922641994345926,
602
+ "grad_norm": 5.369118446973871,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.6399550437927246,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.11051143664867644,
609
+ "grad_norm": 2.44005948506668,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.6399797439575196,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.1117964533538936,
616
+ "grad_norm": 9.032870562997907,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.6228564739227295,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.11308147005911078,
623
+ "grad_norm": 2.2222412773073255,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.6560338497161865,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.11436648676432794,
630
+ "grad_norm": 3.2511137004113686,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.637030029296875,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.1156515034695451,
637
+ "grad_norm": 2.19881020755354,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.6965466499328613,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.11693652017476228,
644
+ "grad_norm": 1.5817637004939435,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.6543152809143067,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.11822153687997944,
651
+ "grad_norm": 1.1276990840531103,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.6320334434509277,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.1195065535851966,
658
+ "grad_norm": 1.5922154872413146,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.617594051361084,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.12079157029041378,
665
+ "grad_norm": 2.504533267963736,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.609941816329956,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.12207658699563094,
672
+ "grad_norm": 1.0654835934253888,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.6268711090087891,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.12336160370084812,
679
+ "grad_norm": 1.3754530051132428,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.6483760833740234,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.12464662040606528,
686
+ "grad_norm": 2.0675680343817304,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.625995397567749,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.12593163711128244,
693
+ "grad_norm": 0.9779567253189434,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.6271099090576172,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.1272166538164996,
700
+ "grad_norm": 0.9982099110220216,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.61366286277771,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.1285016705217168,
707
+ "grad_norm": 1.0658286710551574,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.6421555519104004,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.12978668722693396,
714
+ "grad_norm": 1.2925354348794225,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.6284059524536133,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.13107170393215112,
721
+ "grad_norm": 1.0082444749487007,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.6287686824798584,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.13235672063736828,
728
+ "grad_norm": 1.4045341222523071,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.6289882659912109,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.13364173734258544,
735
+ "grad_norm": 1.1867690266586628,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.6120185852050781,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.13492675404780263,
742
+ "grad_norm": 1.0489340200173105,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.6052242279052734,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.1362117707530198,
749
+ "grad_norm": 1.3799884020710995,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.6165138244628906,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.13749678745823696,
756
+ "grad_norm": 2.9370340336765803,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.6380349159240722,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.13878180416345412,
763
+ "grad_norm": 1.0910366276944916,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.6134048461914062,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.14006682086867128,
770
+ "grad_norm": 1.099783169446652,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.6116578578948975,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.14135183757388847,
777
+ "grad_norm": 6.3506737723576645,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.6087802410125732,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.14263685427910563,
784
+ "grad_norm": 1.830326258724304,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.6111043453216553,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.1439218709843228,
791
+ "grad_norm": 1.4371773206636391,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.6332690238952636,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.14520688768953996,
798
+ "grad_norm": 1.1346896882724038,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.6117291450500488,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.14649190439475712,
805
+ "grad_norm": 1.229043968440211,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.634147834777832,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.1477769210999743,
812
+ "grad_norm": 1.0643282673936258,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.6050711631774902,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.14906193780519147,
819
+ "grad_norm": 0.9899183553255214,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.6159072875976562,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.15034695451040864,
826
+ "grad_norm": 0.9756944895379619,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.6437289714813232,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.1516319712156258,
833
+ "grad_norm": 1.5804530368583096,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.6059542655944824,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.15291698792084296,
840
+ "grad_norm": 2.653732074916257,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.6155671119689942,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.15420200462606015,
847
+ "grad_norm": 1.0526988484175226,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.607187032699585,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.15548702133127731,
854
+ "grad_norm": 1.4066545066883747,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.6067052841186523,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.15677203803649448,
861
+ "grad_norm": 1.14445816579696,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.6031442165374756,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.15805705474171164,
868
+ "grad_norm": 0.9688379604966122,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.6165518760681152,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.1593420714469288,
875
+ "grad_norm": 0.8892837373403698,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.6012338161468506,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.160627088152146,
882
+ "grad_norm": 1.4152027685469248,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.5886805534362793,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.16191210485736315,
889
+ "grad_norm": 2.1457539636692267,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.6081838130950927,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.16319712156258032,
896
+ "grad_norm": 1.7040104665899867,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.6133224487304687,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.16448213826779748,
903
+ "grad_norm": 1.2352798817335287,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.6118340492248535,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.16576715497301464,
910
+ "grad_norm": 1.2198439029279724,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.6088254451751709,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.1670521716782318,
917
+ "grad_norm": 3.6356160898778853,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.5922161102294922,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.168337188383449,
924
+ "grad_norm": 1.434121136195825,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.6113036155700684,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.16962220508866616,
931
+ "grad_norm": 1.8215041680182131,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.6138492584228515,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.17090722179388332,
938
+ "grad_norm": 0.9733584275000338,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.6006131649017334,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.17219223849910048,
945
+ "grad_norm": 0.8291757298058455,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.6175620555877686,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.17347725520431764,
952
+ "grad_norm": 1.7778569183139799,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.6048360347747803,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.17476227190953483,
959
+ "grad_norm": 2.3344157936720307,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.6050002098083496,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.176047288614752,
966
+ "grad_norm": 3.506671656079186,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.6108762741088867,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.17733230531996916,
973
+ "grad_norm": 1.1491594873900661,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.5949950695037842,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.17861732202518632,
980
+ "grad_norm": 0.9698616451139325,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.6035851955413818,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.17990233873040348,
987
+ "grad_norm": 2.3293601794536607,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.6041230201721192,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.18118735543562067,
994
+ "grad_norm": 0.9819292447087378,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.6033549308776855,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.18247237214083784,
1001
+ "grad_norm": 5.127718733826317,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.5804622650146485,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.183757388846055,
1008
+ "grad_norm": 1.3831078732533943,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.5929896354675293,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.18504240555127216,
1015
+ "grad_norm": 1.409835795434802,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.620157527923584,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.18632742225648932,
1022
+ "grad_norm": 0.9787989385784489,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.6010499477386475,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.1876124389617065,
1029
+ "grad_norm": 1.125282063521845,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.6063499450683594,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.18889745566692367,
1036
+ "grad_norm": 1.1878340415798876,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.6011593818664551,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.19018247237214084,
1043
+ "grad_norm": 1.0586030412435297,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.5826192378997803,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.191467489077358,
1050
+ "grad_norm": 1.4030416349158283,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.5979415893554687,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.19275250578257516,
1057
+ "grad_norm": 0.9882974034594119,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.5784321784973144,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.19403752248779235,
1064
+ "grad_norm": 1.5827378342447296,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.5972190380096436,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.19532253919300951,
1071
+ "grad_norm": 1.772835353002146,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.5883693218231201,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.19660755589822668,
1078
+ "grad_norm": 0.9504309854461958,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.5878360748291016,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.19789257260344384,
1085
+ "grad_norm": 3.4573479282773754,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.5961189270019531,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.199177589308661,
1092
+ "grad_norm": 1.068378062020343,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.6108866691589355,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.2004626060138782,
1099
+ "grad_norm": 0.9475492216550739,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.5879310607910156,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.20174762271909535,
1106
+ "grad_norm": 2.6996846803236796,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.6067344665527343,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.20303263942431252,
1113
+ "grad_norm": 1.503312484489778,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.5966094970703125,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.20431765612952968,
1120
+ "grad_norm": 2.1144705955722105,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.601826810836792,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.20560267283474684,
1127
+ "grad_norm": 2.0298939231075686,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.613339900970459,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.20688768953996403,
1134
+ "grad_norm": 0.9838666275922071,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.5876136779785156,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.2081727062451812,
1141
+ "grad_norm": 1.3776226412627304,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.5954419612884522,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.20945772295039836,
1148
+ "grad_norm": 0.8085782872052595,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.595486831665039,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.21074273965561552,
1155
+ "grad_norm": 0.8662216027678745,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.5841521263122559,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.21202775636083268,
1162
+ "grad_norm": 1.2239187193929486,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.5863484859466552,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.21331277306604987,
1169
+ "grad_norm": 1.0494852911396721,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.5841124534606934,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.21459778977126703,
1176
+ "grad_norm": 2.981594663490346,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.5967587947845459,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.2158828064764842,
1183
+ "grad_norm": 1.296867754625597,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.6040875434875488,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.21716782318170136,
1190
+ "grad_norm": 1.6322116014400065,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.6083320617675781,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.21845283988691852,
1197
+ "grad_norm": 1.1441615552670734,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.6085511684417725,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.2197378565921357,
1204
+ "grad_norm": 0.9746942510805159,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.5959947586059571,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.22102287329735287,
1211
+ "grad_norm": 1.847667452156604,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.5838113307952881,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.22230789000257004,
1218
+ "grad_norm": 1.3014028403342264,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.5898797988891602,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.2235929067077872,
1225
+ "grad_norm": 1.794520627282578,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.5770421504974366,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.22487792341300436,
1232
+ "grad_norm": 1.181651271225444,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.5804369926452637,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.22616294011822155,
1239
+ "grad_norm": 1.4302258126941458,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.5999673843383789,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.2274479568234387,
1246
+ "grad_norm": 1.5184617174292843,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.601544713973999,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.22873297352865588,
1253
+ "grad_norm": 1.3090615344632057,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.5916304588317871,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.23001799023387304,
1260
+ "grad_norm": 1.084083761154048,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.599252462387085,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.2313030069390902,
1267
+ "grad_norm": 1.06362134663052,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.5892449855804444,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.23258802364430736,
1274
+ "grad_norm": 1.3218270041853413,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.5810562133789062,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.23387304034952455,
1281
+ "grad_norm": 1.8735021692385667,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.563934326171875,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.23515805705474171,
1288
+ "grad_norm": 1.1029961980816403,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.6061427116394043,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.23644307375995888,
1295
+ "grad_norm": 1.0956401424009607,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.580710506439209,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.23772809046517604,
1302
+ "grad_norm": 0.9022890666227324,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.5831015586853028,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.2390131071703932,
1309
+ "grad_norm": 0.9739413408546281,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.5803080558776855,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.2402981238756104,
1316
+ "grad_norm": 0.8887115233132612,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.5808173656463623,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.24158314058082755,
1323
+ "grad_norm": 1.2281390991341632,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.584747314453125,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.24286815728604472,
1330
+ "grad_norm": 1.0293295686123922,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.5973689079284668,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.24415317399126188,
1337
+ "grad_norm": 2.240993859946262,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.5756341934204101,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.24543819069647904,
1344
+ "grad_norm": 2.542763529237098,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.568706464767456,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.24672320740169623,
1351
+ "grad_norm": 1.8658118445309126,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.5815502166748047,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.2480082241069134,
1358
+ "grad_norm": 1.3680858039002373,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.5713666915893555,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.24929324081213056,
1365
+ "grad_norm": 1.4545423045311263,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.5888118267059326,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.25057825751734775,
1372
+ "grad_norm": 1.0001999301868607,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.581107234954834,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.2518632742225649,
1379
+ "grad_norm": 3.804939278531223,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.5979499816894531,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.25314829092778207,
1386
+ "grad_norm": 1.6052651400683768,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.5852889060974121,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.2544333076329992,
1393
+ "grad_norm": 1.465455399606115,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.5984594345092773,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.2557183243382164,
1400
+ "grad_norm": 0.9532879108802271,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.5933744430541992,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.2570033410434336,
1407
+ "grad_norm": 0.9971702008546999,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.5816676616668701,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.2582883577486507,
1414
+ "grad_norm": 0.9234236488287841,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.5809596538543701,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.2595733744538679,
1421
+ "grad_norm": 1.2689341107385237,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.5777791023254395,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.26085839115908505,
1428
+ "grad_norm": 2.127441593187859,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.5782732963562012,
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 sha256:4842e6480a0855b865ad1660ab987ba6262c9c14a138579b35569eb5338975ee
3
+ size 7313