Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-g2_1_oldcode_cm_newcode

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
894e45e
·
verified ·
1 Parent(s): a46559a

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:cd283ad0593ae07a0d4380550bd8eb6baa792dbcb520da49c9c30d0142758858
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": 0.03970499287846565,
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": 0.0001951105301153103,
14
+ "grad_norm": 1290.1151800966732,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 1.6666240692138672,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0003902210602306206,
21
+ "grad_norm": 2094.709006858091,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 0.9677549362182617,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.000585331590345931,
28
+ "grad_norm": 30.714148238594614,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.7858523368835449,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.0007804421204612412,
35
+ "grad_norm": 497.9186820845375,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.7518716812133789,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.0009755526505765516,
42
+ "grad_norm": 40.68707475383215,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.7165522575378418,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.001170663180691862,
49
+ "grad_norm": 49.16345817267902,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.6980477333068847,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0013657737108071723,
56
+ "grad_norm": 110.36540709084935,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.6940877437591553,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.0015608842409224825,
63
+ "grad_norm": 45.22696804420163,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.7629052639007569,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.001755994771037793,
70
+ "grad_norm": 8.589551842197837,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.7278446197509766,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.0019511053011531032,
77
+ "grad_norm": 22.256653345446292,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.7335631370544433,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.0021462158312684134,
84
+ "grad_norm": 37.8295200972987,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.7461274623870849,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.002341326361383724,
91
+ "grad_norm": 15.77833948739335,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.7302251338958741,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.0025364368914990343,
98
+ "grad_norm": 662.5019348463974,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.7099417686462403,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.0027315474216143445,
105
+ "grad_norm": 8.07873176220526,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 0.8300569534301758,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.0029266579517296548,
112
+ "grad_norm": 108.63732872651582,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.7591472148895264,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.003121768481844965,
119
+ "grad_norm": 12.742460468210709,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.7846505641937256,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.0033168790119602757,
126
+ "grad_norm": 19.46685500550541,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.7010777473449707,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.003511989542075586,
133
+ "grad_norm": 3.79492579495788,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.7221960067749024,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.003707100072190896,
140
+ "grad_norm": 3.761910629826001,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.7448129653930664,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.0039022106023062064,
147
+ "grad_norm": 4.471289857736554,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.7120450496673584,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.004097321132421517,
154
+ "grad_norm": 7.832558161000839,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.7076560974121093,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.004292431662536827,
161
+ "grad_norm": 24.625069340034873,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.7173271179199219,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.004487542192652137,
168
+ "grad_norm": 10.643240144364263,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.8566647529602051,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.004682652722767448,
175
+ "grad_norm": 17.191035538928677,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.7157638549804688,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.004877763252882758,
182
+ "grad_norm": 13.627404303033932,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.7801154136657715,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.005072873782998069,
189
+ "grad_norm": 24.99914666002505,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.7654953002929688,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.005267984313113379,
196
+ "grad_norm": 147.25366832545367,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.7404791831970214,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.005463094843228689,
203
+ "grad_norm": 6.625240044249601,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.7907636165618896,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.005658205373343999,
210
+ "grad_norm": 11.741773313405414,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.7048698425292969,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.0058533159034593095,
217
+ "grad_norm": 5.519224801929464,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.7155918121337891,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.00604842643357462,
224
+ "grad_norm": 11.598372302819296,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.7381957054138184,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.00624353696368993,
231
+ "grad_norm": 9.444939061369409,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.7207625865936279,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.006438647493805241,
238
+ "grad_norm": 7.801173531822594,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.723687744140625,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.006633758023920551,
245
+ "grad_norm": 1.9811128361432186,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.7044608116149902,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.0068288685540358616,
252
+ "grad_norm": 5.818968297777697,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.7668921947479248,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.007023979084151172,
259
+ "grad_norm": 3.944741659070438,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.7118455410003662,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.007219089614266482,
266
+ "grad_norm": 4.311433759305494,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.7700147151947021,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.007414200144381792,
273
+ "grad_norm": 30.425032251061474,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.7222751140594482,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.0076093106744971025,
280
+ "grad_norm": 2.782539467868993,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.6802415370941162,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.007804421204612413,
287
+ "grad_norm": 6.499874836624361,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.6992295742034912,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.007999531734727724,
294
+ "grad_norm": 5.632464299732364,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.7049863815307618,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.008194642264843033,
301
+ "grad_norm": 11.183347233038614,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.7416215419769288,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.008389752794958344,
308
+ "grad_norm": 7.528401125738135,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.7592568397521973,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.008584863325073654,
315
+ "grad_norm": 5.002885634433326,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.756533145904541,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.008779973855188965,
322
+ "grad_norm": 5.66603343994339,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.7430764198303222,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.008975084385304274,
329
+ "grad_norm": 22.314858498949487,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.7967084884643555,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.009170194915419585,
336
+ "grad_norm": 5.642840192700939,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.7398099422454834,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.009365305445534896,
343
+ "grad_norm": 2.0656284529162243,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.7248655319213867,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.009560415975650206,
350
+ "grad_norm": 4.657301334726203,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.7701823711395264,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.009755526505765517,
357
+ "grad_norm": 4.995971591382155,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.733971357345581,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.009950637035880826,
364
+ "grad_norm": 11.657412731021536,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.802979564666748,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.010145747565996137,
371
+ "grad_norm": 3.161728049215805,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.753895378112793,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.010340858096111447,
378
+ "grad_norm": 24.0896519055277,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.7964776992797852,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.010535968626226758,
385
+ "grad_norm": 8.361986893202518,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.7574877738952637,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.010731079156342067,
392
+ "grad_norm": 10.53310762606766,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.7854220867156982,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.010926189686457378,
399
+ "grad_norm": 96.65480630904233,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.7968769073486328,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.01112130021657269,
406
+ "grad_norm": 1.8376056131220815,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.7898524284362793,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.011316410746687999,
413
+ "grad_norm": 1.446906475407358,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.7536652565002442,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.01151152127680331,
420
+ "grad_norm": 2.63100640718635,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.7614954948425293,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.011706631806918619,
427
+ "grad_norm": 1.7292424310123347,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.6910699844360352,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.01190174233703393,
434
+ "grad_norm": 1.6185308860071268,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.7510101318359375,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.01209685286714924,
441
+ "grad_norm": 9.938109528270555,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.7183078765869141,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.01229196339726455,
448
+ "grad_norm": 3.1451531777998762,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.6925305843353271,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.01248707392737986,
455
+ "grad_norm": 11.976814414360483,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.6874044418334961,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.012682184457495171,
462
+ "grad_norm": 2.512055129296439,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.7078936576843262,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.012877294987610482,
469
+ "grad_norm": 21.321440055444356,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.7755936622619629,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.013072405517725792,
476
+ "grad_norm": 6.027137898134079,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.6783116817474365,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.013267516047841103,
483
+ "grad_norm": 1.6349976601766352,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.7179695129394531,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.013462626577956412,
490
+ "grad_norm": 2.696759389003869,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.6627010345458985,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.013657737108071723,
497
+ "grad_norm": 3.094531262148124,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.691745662689209,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.013852847638187032,
504
+ "grad_norm": 1.3167065811312815,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.6898958206176757,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.014047958168302344,
511
+ "grad_norm": 3.096688041091074,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.6983632564544677,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.014243068698417653,
518
+ "grad_norm": 1.5732201392921932,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.6881768703460693,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.014438179228532964,
525
+ "grad_norm": 4.179444784177794,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.6782029628753662,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.014633289758648273,
532
+ "grad_norm": 13.022740345020917,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.7157995223999023,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.014828400288763584,
539
+ "grad_norm": 2.227508854051503,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.7245219230651856,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.015023510818878896,
546
+ "grad_norm": 3.4074041057703774,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.7473674297332764,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.015218621348994205,
553
+ "grad_norm": 4.053610258594328,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.664447021484375,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.015413731879109516,
560
+ "grad_norm": 1.6609484317874577,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.6886637687683106,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.015608842409224825,
567
+ "grad_norm": 1.5444399305626002,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.7088738918304444,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.015803952939340137,
574
+ "grad_norm": 3.1762223604517237,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.7194277763366699,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.015999063469455448,
581
+ "grad_norm": 3.8350609994341527,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.7160842895507813,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.016194173999570755,
588
+ "grad_norm": 2.315457264069663,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.6816580772399903,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.016389284529686066,
595
+ "grad_norm": 3.1340139053276896,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.6640499114990235,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.016584395059801377,
602
+ "grad_norm": 2.103288119861107,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.6783723354339599,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.01677950558991669,
609
+ "grad_norm": 3.7306310450926383,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.6990603446960449,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.016974616120032,
616
+ "grad_norm": 2.3246559374347955,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.6684437751770019,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.017169726650147307,
623
+ "grad_norm": 2.4616580742100016,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.6759738922119141,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.01736483718026262,
630
+ "grad_norm": 1.4211835502848627,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.7110819816589355,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.01755994771037793,
637
+ "grad_norm": 3.113362922049139,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.7206232070922851,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.01775505824049324,
644
+ "grad_norm": 1.5202175869081638,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.6578310012817383,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.017950168770608548,
651
+ "grad_norm": 1.610151123997545,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.7033986568450927,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.01814527930072386,
658
+ "grad_norm": 2.213950402852014,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.7181561946868896,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.01834038983083917,
665
+ "grad_norm": 1.0525256653678736,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.64858980178833,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.01853550036095448,
672
+ "grad_norm": 1.1535760888310356,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.6768205642700196,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.018730610891069793,
679
+ "grad_norm": 2.5687267328441172,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.7187290191650391,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.0189257214211851,
686
+ "grad_norm": 2.363644647745971,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.6914881706237793,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.01912083195130041,
693
+ "grad_norm": 1.6586735289269663,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.7053565979003906,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.019315942481415722,
700
+ "grad_norm": 1.3154384327675313,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.6894444942474365,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.019511053011531033,
707
+ "grad_norm": 1.4829582302075024,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.6964797019958496,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.01970616354164634,
714
+ "grad_norm": 11.246853734884391,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.6903947830200196,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.019901274071761652,
721
+ "grad_norm": 2.130270630304158,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.7131201744079589,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.020096384601876963,
728
+ "grad_norm": 2.3843191059461684,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.7245532989501953,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.020291495131992274,
735
+ "grad_norm": 1.8036658748418921,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.688844633102417,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.020486605662107586,
742
+ "grad_norm": 4.142854989409903,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.6744081020355225,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.020681716192222893,
749
+ "grad_norm": 1.472431946736185,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.6328347682952881,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.020876826722338204,
756
+ "grad_norm": 2.390747409072109,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.688608455657959,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.021071937252453515,
763
+ "grad_norm": 2.538662996092993,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.7029477119445801,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.021267047782568826,
770
+ "grad_norm": 2.4656407868003436,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.6541564464569092,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.021462158312684134,
777
+ "grad_norm": 1.4483652415733586,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.6665634155273438,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.021657268842799445,
784
+ "grad_norm": 3.7600001493285813,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.6685405731201172,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.021852379372914756,
791
+ "grad_norm": 4.4698831654409785,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.6686572074890137,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.022047489903030067,
798
+ "grad_norm": 1.2416499167615647,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.6792525291442871,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.02224260043314538,
805
+ "grad_norm": 1.261439766703017,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.6553254127502441,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.022437710963260686,
812
+ "grad_norm": 3.0580767819341967,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.670100736618042,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.022632821493375997,
819
+ "grad_norm": 1.9982728200546733,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.6652833461761475,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.02282793202349131,
826
+ "grad_norm": 1.4478470019999148,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.6802016258239746,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.02302304255360662,
833
+ "grad_norm": 1.390262972826178,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.6520106792449951,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.023218153083721927,
840
+ "grad_norm": 1.2517531765951326,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.6596714973449707,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.023413263613837238,
847
+ "grad_norm": 1.66285797208597,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.6890499114990234,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.02360837414395255,
854
+ "grad_norm": 1.4023367573944592,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.6909699440002441,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.02380348467406786,
861
+ "grad_norm": 2.103600383565939,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.6514087677001953,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.02399859520418317,
868
+ "grad_norm": 1.3538205014153204,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.7010597229003906,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.02419370573429848,
875
+ "grad_norm": 1.131100957434826,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.6904762268066407,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.02438881626441379,
882
+ "grad_norm": 0.8521222417760329,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.6021539688110351,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.0245839267945291,
889
+ "grad_norm": 0.9985116086981741,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.6618970394134521,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.024779037324644412,
896
+ "grad_norm": 1.0359703256994235,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.6823967933654785,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.02497414785475972,
903
+ "grad_norm": 1.2887090486112476,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.6628634452819824,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.02516925838487503,
910
+ "grad_norm": 3.432834498196539,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.7419169425964356,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.025364368914990342,
917
+ "grad_norm": 23.557815006813602,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.7083474636077881,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.025559479445105653,
924
+ "grad_norm": 1.3779699418257583,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.6835464477539063,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.025754589975220964,
931
+ "grad_norm": 1.0971280641281729,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.6729559898376465,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.025949700505336272,
938
+ "grad_norm": 86.50430124250552,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.7425766944885254,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.026144811035451583,
945
+ "grad_norm": 1.9060666983098362,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.6884101867675781,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.026339921565566894,
952
+ "grad_norm": 4.952310768605313,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.6712810516357421,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.026535032095682205,
959
+ "grad_norm": 2.1836902499589055,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.6641899108886719,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.026730142625797513,
966
+ "grad_norm": 0.9745030121353373,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.6548121929168701,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.026925253155912824,
973
+ "grad_norm": 1.0085691573367601,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.6872680187225342,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.027120363686028135,
980
+ "grad_norm": 1.0328562743257255,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.6593247413635254,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.027315474216143446,
987
+ "grad_norm": 1.7706110276938303,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.6563483715057373,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.027510584746258757,
994
+ "grad_norm": 1.1475044260212954,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.6485116004943847,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.027705695276374065,
1001
+ "grad_norm": 2.1343320521467724,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.6518092155456543,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.027900805806489376,
1008
+ "grad_norm": 19.40190752512825,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.6708407402038574,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.028095916336604687,
1015
+ "grad_norm": 1.0429828458514099,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.6531154632568359,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.028291026866719998,
1022
+ "grad_norm": 3.854687583398717,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.6318797588348388,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.028486137396835306,
1029
+ "grad_norm": 1.2803239369157802,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.6483430385589599,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.028681247926950617,
1036
+ "grad_norm": 1.7844005133176821,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.670600414276123,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.028876358457065928,
1043
+ "grad_norm": 2.638360353786163,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.6612592220306397,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.02907146898718124,
1050
+ "grad_norm": 1.2239072376785438,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.6572507381439209,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.029266579517296547,
1057
+ "grad_norm": 3.9817061848469457,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.631273603439331,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.029461690047411858,
1064
+ "grad_norm": 1.2083656296744134,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.6754149913787841,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.02965680057752717,
1071
+ "grad_norm": 1.6701538309486328,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.6359077453613281,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.02985191110764248,
1078
+ "grad_norm": 1.1088864730268388,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.6407184600830078,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.03004702163775779,
1085
+ "grad_norm": 2.855041461691364,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.6601726531982421,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.0302421321678731,
1092
+ "grad_norm": 1.041676019536129,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.6693766593933106,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.03043724269798841,
1099
+ "grad_norm": 1.7903036215917778,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.6842814445495605,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.03063235322810372,
1106
+ "grad_norm": 1.2275140901739492,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.6563383102416992,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.030827463758219032,
1113
+ "grad_norm": 1.4416525013205308,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.70062255859375,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.03102257428833434,
1120
+ "grad_norm": 1.022694275409909,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.6751978874206543,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.03121768481844965,
1127
+ "grad_norm": 1.5434339515015798,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.6534595489501953,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.03141279534856496,
1134
+ "grad_norm": 4.482223058250825,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.679110336303711,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.03160790587868027,
1141
+ "grad_norm": 1.269130777338052,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.6603041648864746,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.031803016408795584,
1148
+ "grad_norm": 3.5729377498646477,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.6539816856384277,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.031998126938910895,
1155
+ "grad_norm": 2.6023274014566247,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.7010405540466309,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.032193237469026206,
1162
+ "grad_norm": 9.210714854982013,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.6686701774597168,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.03238834799914151,
1169
+ "grad_norm": 5.165666638947227,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.6095908164978028,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.03258345852925682,
1176
+ "grad_norm": 9.16141664239496,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.7108460903167725,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.03277856905937213,
1183
+ "grad_norm": 3.161658879203255,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.6284823417663574,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.032973679589487444,
1190
+ "grad_norm": 5.944141446716445,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.6405398845672607,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.033168790119602755,
1197
+ "grad_norm": 1.7808423949894736,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.6491862297058105,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.033363900649718066,
1204
+ "grad_norm": 1.726175898675113,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.6724005699157715,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.03355901117983338,
1211
+ "grad_norm": 2.376245899157911,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.6525126934051514,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.03375412170994869,
1218
+ "grad_norm": 1.6653879954838984,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.6367118835449219,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.033949232240064,
1225
+ "grad_norm": 1.3883560501158865,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.6553079128265381,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.0341443427701793,
1232
+ "grad_norm": 11.844497956339469,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.5972353935241699,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.034339453300294615,
1239
+ "grad_norm": 1.5888196717274026,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.658538818359375,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.034534563830409926,
1246
+ "grad_norm": 1.2587083645893846,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.6199719905853271,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.03472967436052524,
1253
+ "grad_norm": 2.0204514597287266,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.6569772243499756,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.03492478489064055,
1260
+ "grad_norm": 1.1323396524728921,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.6597754478454589,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.03511989542075586,
1267
+ "grad_norm": 4.803084432276915,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.6832777976989746,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.03531500595087117,
1274
+ "grad_norm": 1.2706249626426194,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.6612133026123047,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.03551011648098648,
1281
+ "grad_norm": 2.888385667334179,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.6800965785980224,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.03570522701110179,
1288
+ "grad_norm": 1.7994196905308342,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.6731705665588379,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.035900337541217096,
1295
+ "grad_norm": 2.2229464990180463,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.6999088287353515,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.03609544807133241,
1302
+ "grad_norm": 1.7473284581429838,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.665509557723999,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.03629055860144772,
1309
+ "grad_norm": 2.2975536238797036,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.6647231101989746,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.03648566913156303,
1316
+ "grad_norm": 2.6214877289682246,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.62739577293396,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.03668077966167834,
1323
+ "grad_norm": 2.192373938003488,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.7087226390838623,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.03687589019179365,
1330
+ "grad_norm": 6.62847066861488,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.6841737270355225,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.03707100072190896,
1337
+ "grad_norm": 8.573387218725129,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.6550696849822998,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.037266111252024274,
1344
+ "grad_norm": 21.81644893450406,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.6534019470214844,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.037461221782139585,
1351
+ "grad_norm": 1.972315212149555,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.6699990272521973,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.03765633231225489,
1358
+ "grad_norm": 1.572875023378223,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.6702420234680175,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.0378514428423702,
1365
+ "grad_norm": 1.2077505710786207,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.6410272121429443,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.03804655337248551,
1372
+ "grad_norm": 1.3936237416130755,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.6933211803436279,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.03824166390260082,
1379
+ "grad_norm": 2.6969645830470323,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.6899444580078125,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.038436774432716134,
1386
+ "grad_norm": 3.6883874679320745,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.661356258392334,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.038631884962831445,
1393
+ "grad_norm": 4.1375939813313325,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.6723477840423584,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.038826995492946756,
1400
+ "grad_norm": 8.712693611376435,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.6563708305358886,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.03902210602306207,
1407
+ "grad_norm": 1.7546049733134041,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.6711672782897949,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.03921721655317738,
1414
+ "grad_norm": 1.7813893766887805,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.6956442832946778,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.03941232708329268,
1421
+ "grad_norm": 1.2906449372308841,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.6290652275085449,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.03960743761340799,
1428
+ "grad_norm": 1.1252587686982205,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.6268731117248535,
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 sha256:70e4e3435ce903e96f0f05ad12e0be3c1bc36afff496aa175c867be51fe32316
3
+ size 7313