Training in progress, step 12816
Browse files- checkpoint-12816/config.json +39 -0
- checkpoint-12816/merges.txt +0 -0
- checkpoint-12816/optimizer.pt +3 -0
- checkpoint-12816/pytorch_model.bin +3 -0
- checkpoint-12816/rng_state.pth +3 -0
- checkpoint-12816/scaler.pt +3 -0
- checkpoint-12816/scheduler.pt +3 -0
- checkpoint-12816/special_tokens_map.json +6 -0
- checkpoint-12816/tokenizer.json +0 -0
- checkpoint-12816/tokenizer_config.json +10 -0
- checkpoint-12816/trainer_state.json +1884 -0
- checkpoint-12816/training_args.bin +3 -0
- checkpoint-12816/vocab.json +0 -0
- pytorch_model.bin +1 -1
checkpoint-12816/config.json
ADDED
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "gpt2",
|
3 |
+
"activation_function": "gelu_new",
|
4 |
+
"architectures": [
|
5 |
+
"GPT2LMAndValueHeadModel"
|
6 |
+
],
|
7 |
+
"attn_pdrop": 0.1,
|
8 |
+
"bos_token_id": 50256,
|
9 |
+
"embd_pdrop": 0.1,
|
10 |
+
"eos_token_id": 50256,
|
11 |
+
"initializer_range": 0.02,
|
12 |
+
"layer_norm_epsilon": 1e-05,
|
13 |
+
"model_type": "gpt2",
|
14 |
+
"n_ctx": 1024,
|
15 |
+
"n_embd": 768,
|
16 |
+
"n_head": 12,
|
17 |
+
"n_inner": null,
|
18 |
+
"n_layer": 12,
|
19 |
+
"n_positions": 1024,
|
20 |
+
"reorder_and_upcast_attn": true,
|
21 |
+
"resid_pdrop": 0.1,
|
22 |
+
"scale_attn_by_inverse_layer_idx": false,
|
23 |
+
"scale_attn_weights": true,
|
24 |
+
"summary_activation": null,
|
25 |
+
"summary_first_dropout": 0.1,
|
26 |
+
"summary_proj_to_labels": true,
|
27 |
+
"summary_type": "cls_index",
|
28 |
+
"summary_use_proj": true,
|
29 |
+
"task_specific_params": {
|
30 |
+
"text-generation": {
|
31 |
+
"do_sample": true,
|
32 |
+
"max_length": 50
|
33 |
+
}
|
34 |
+
},
|
35 |
+
"torch_dtype": "float32",
|
36 |
+
"transformers_version": "4.23.0",
|
37 |
+
"use_cache": true,
|
38 |
+
"vocab_size": 50257
|
39 |
+
}
|
checkpoint-12816/merges.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-12816/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c5af1ef8682549010f77d66558facbc46d0c139c1a7b9e5a9614c965479f15fe
|
3 |
+
size 995605189
|
checkpoint-12816/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3bc4cf094124a3fd9b78471cdeec9c812040347ceebfa4c5c137d8a18ca6c37d
|
3 |
+
size 510398013
|
checkpoint-12816/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ab7e9975f1b8bae022fb24372df86aa196532d115d6fefc3227fc58eb4dbd9d4
|
3 |
+
size 15597
|
checkpoint-12816/scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6da637ba916ce0c18260778dc23c88f0cbebfb69369dbb3893b2d1912f0651bf
|
3 |
+
size 557
|
checkpoint-12816/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f0dbe3b5bfa7e83f85944afc5d55068d08e05ef159701e9d374c9dc11db264ec
|
3 |
+
size 627
|
checkpoint-12816/special_tokens_map.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<|endoftext|>",
|
3 |
+
"eos_token": "<|endoftext|>",
|
4 |
+
"pad_token": "<|endoftext|>",
|
5 |
+
"unk_token": "<|endoftext|>"
|
6 |
+
}
|
checkpoint-12816/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-12816/tokenizer_config.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_prefix_space": false,
|
3 |
+
"bos_token": "<|endoftext|>",
|
4 |
+
"eos_token": "<|endoftext|>",
|
5 |
+
"model_max_length": 1024,
|
6 |
+
"name_or_path": "gpt2",
|
7 |
+
"special_tokens_map_file": null,
|
8 |
+
"tokenizer_class": "GPT2Tokenizer",
|
9 |
+
"unk_token": "<|endoftext|>"
|
10 |
+
}
|
checkpoint-12816/trainer_state.json
ADDED
@@ -0,0 +1,1884 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.2999719127422526,
|
5 |
+
"global_step": 12816,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.0,
|
12 |
+
"learning_rate": 1.1682242990654204e-06,
|
13 |
+
"loss": 10.9364,
|
14 |
+
"theoretical_loss": 20.81281780154715,
|
15 |
+
"tokens_seen": 65536
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.0,
|
19 |
+
"learning_rate": 5.841121495327103e-05,
|
20 |
+
"loss": 8.9947,
|
21 |
+
"theoretical_loss": 8.563482664611069,
|
22 |
+
"tokens_seen": 3276800
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"epoch": 0.0,
|
26 |
+
"learning_rate": 0.00011682242990654206,
|
27 |
+
"loss": 7.0133,
|
28 |
+
"theoretical_loss": 7.4777587180480305,
|
29 |
+
"tokens_seen": 6553600
|
30 |
+
},
|
31 |
+
{
|
32 |
+
"epoch": 0.0,
|
33 |
+
"learning_rate": 0.00017523364485981307,
|
34 |
+
"loss": 6.2244,
|
35 |
+
"theoretical_loss": 6.9337544888949,
|
36 |
+
"tokens_seen": 9830400
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"epoch": 0.0,
|
40 |
+
"learning_rate": 0.00023364485981308412,
|
41 |
+
"loss": 5.822,
|
42 |
+
"theoretical_loss": 6.583566228426414,
|
43 |
+
"tokens_seen": 13107200
|
44 |
+
},
|
45 |
+
{
|
46 |
+
"epoch": 0.01,
|
47 |
+
"learning_rate": 0.00029205607476635517,
|
48 |
+
"loss": 5.5642,
|
49 |
+
"theoretical_loss": 6.330713565116083,
|
50 |
+
"tokens_seen": 16384000
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.01,
|
54 |
+
"learning_rate": 0.00035046728971962614,
|
55 |
+
"loss": 5.3688,
|
56 |
+
"theoretical_loss": 6.135529231940326,
|
57 |
+
"tokens_seen": 19660800
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.01,
|
61 |
+
"learning_rate": 0.0004088785046728972,
|
62 |
+
"loss": 5.1987,
|
63 |
+
"theoretical_loss": 5.978101583869607,
|
64 |
+
"tokens_seen": 22937600
|
65 |
+
},
|
66 |
+
{
|
67 |
+
"epoch": 0.01,
|
68 |
+
"learning_rate": 0.00046728971962616824,
|
69 |
+
"loss": 5.1026,
|
70 |
+
"theoretical_loss": 5.8471173262659235,
|
71 |
+
"tokens_seen": 26214400
|
72 |
+
},
|
73 |
+
{
|
74 |
+
"epoch": 0.01,
|
75 |
+
"learning_rate": 0.000499739928125591,
|
76 |
+
"loss": 4.9312,
|
77 |
+
"theoretical_loss": 5.7355768158821245,
|
78 |
+
"tokens_seen": 29491200
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"epoch": 0.01,
|
82 |
+
"learning_rate": 0.0004991488556837526,
|
83 |
+
"loss": 4.8598,
|
84 |
+
"theoretical_loss": 5.638870144071353,
|
85 |
+
"tokens_seen": 32768000
|
86 |
+
},
|
87 |
+
{
|
88 |
+
"epoch": 0.01,
|
89 |
+
"learning_rate": 0.0004985577832419141,
|
90 |
+
"loss": 4.7297,
|
91 |
+
"theoretical_loss": 5.553812381844907,
|
92 |
+
"tokens_seen": 36044800
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 0.01,
|
96 |
+
"learning_rate": 0.0004979667108000757,
|
97 |
+
"loss": 4.6466,
|
98 |
+
"theoretical_loss": 5.478118080556438,
|
99 |
+
"tokens_seen": 39321600
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.02,
|
103 |
+
"learning_rate": 0.0004973756383582371,
|
104 |
+
"loss": 4.5646,
|
105 |
+
"theoretical_loss": 5.410095959579362,
|
106 |
+
"tokens_seen": 42598400
|
107 |
+
},
|
108 |
+
{
|
109 |
+
"epoch": 0.02,
|
110 |
+
"learning_rate": 0.0004967845659163987,
|
111 |
+
"loss": 4.5162,
|
112 |
+
"theoretical_loss": 5.348462083735834,
|
113 |
+
"tokens_seen": 45875200
|
114 |
+
},
|
115 |
+
{
|
116 |
+
"epoch": 0.02,
|
117 |
+
"learning_rate": 0.0004961934934745603,
|
118 |
+
"loss": 4.3809,
|
119 |
+
"theoretical_loss": 5.292220566937567,
|
120 |
+
"tokens_seen": 49152000
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 0.02,
|
124 |
+
"learning_rate": 0.0004956024210327218,
|
125 |
+
"loss": 4.336,
|
126 |
+
"theoretical_loss": 5.240584625769978,
|
127 |
+
"tokens_seen": 52428800
|
128 |
+
},
|
129 |
+
{
|
130 |
+
"epoch": 0.02,
|
131 |
+
"learning_rate": 0.0004950113485908833,
|
132 |
+
"loss": 4.2829,
|
133 |
+
"theoretical_loss": 5.192922724525789,
|
134 |
+
"tokens_seen": 55705600
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 0.02,
|
138 |
+
"learning_rate": 0.0004944202761490448,
|
139 |
+
"loss": 4.209,
|
140 |
+
"theoretical_loss": 5.1487208633564405,
|
141 |
+
"tokens_seen": 58982400
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 0.02,
|
145 |
+
"learning_rate": 0.0004938292037072064,
|
146 |
+
"loss": 4.0751,
|
147 |
+
"theoretical_loss": 5.107555562405102,
|
148 |
+
"tokens_seen": 62259200
|
149 |
+
},
|
150 |
+
{
|
151 |
+
"epoch": 0.02,
|
152 |
+
"learning_rate": 0.0004932381312653678,
|
153 |
+
"loss": 3.9696,
|
154 |
+
"theoretical_loss": 5.069074117143246,
|
155 |
+
"tokens_seen": 65536000
|
156 |
+
},
|
157 |
+
{
|
158 |
+
"epoch": 0.02,
|
159 |
+
"learning_rate": 0.0004926470588235294,
|
160 |
+
"loss": 3.9197,
|
161 |
+
"theoretical_loss": 5.032979909838007,
|
162 |
+
"tokens_seen": 68812800
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 0.03,
|
166 |
+
"learning_rate": 0.000492055986381691,
|
167 |
+
"loss": 3.8378,
|
168 |
+
"theoretical_loss": 4.999021308224664,
|
169 |
+
"tokens_seen": 72089600
|
170 |
+
},
|
171 |
+
{
|
172 |
+
"epoch": 0.03,
|
173 |
+
"learning_rate": 0.0004914649139398525,
|
174 |
+
"loss": 3.7856,
|
175 |
+
"theoretical_loss": 4.966983155351962,
|
176 |
+
"tokens_seen": 75366400
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 0.03,
|
180 |
+
"learning_rate": 0.000490873841498014,
|
181 |
+
"loss": 3.6988,
|
182 |
+
"theoretical_loss": 4.9366801616251355,
|
183 |
+
"tokens_seen": 78643200
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 0.03,
|
187 |
+
"learning_rate": 0.0004902827690561755,
|
188 |
+
"loss": 3.6826,
|
189 |
+
"theoretical_loss": 4.907951713830082,
|
190 |
+
"tokens_seen": 81920000
|
191 |
+
},
|
192 |
+
{
|
193 |
+
"epoch": 0.03,
|
194 |
+
"learning_rate": 0.0004896916966143371,
|
195 |
+
"loss": 3.6271,
|
196 |
+
"theoretical_loss": 4.880657753812926,
|
197 |
+
"tokens_seen": 85196800
|
198 |
+
},
|
199 |
+
{
|
200 |
+
"epoch": 0.03,
|
201 |
+
"learning_rate": 0.0004891006241724985,
|
202 |
+
"loss": 3.617,
|
203 |
+
"theoretical_loss": 4.854675474481779,
|
204 |
+
"tokens_seen": 88473600
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 0.03,
|
208 |
+
"learning_rate": 0.0004885095517306601,
|
209 |
+
"loss": 3.528,
|
210 |
+
"theoretical_loss": 4.8298966473088125,
|
211 |
+
"tokens_seen": 91750400
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"epoch": 0.03,
|
215 |
+
"learning_rate": 0.0004879184792888217,
|
216 |
+
"loss": 3.5513,
|
217 |
+
"theoretical_loss": 4.8062254427779205,
|
218 |
+
"tokens_seen": 95027200
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 0.04,
|
222 |
+
"learning_rate": 0.0004873274068469832,
|
223 |
+
"loss": 3.4962,
|
224 |
+
"theoretical_loss": 4.783576639276257,
|
225 |
+
"tokens_seen": 98304000
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 0.04,
|
229 |
+
"learning_rate": 0.00048673633440514467,
|
230 |
+
"loss": 3.5334,
|
231 |
+
"theoretical_loss": 4.761874140772408,
|
232 |
+
"tokens_seen": 101580800
|
233 |
+
},
|
234 |
+
{
|
235 |
+
"epoch": 0.04,
|
236 |
+
"learning_rate": 0.0004861452619633062,
|
237 |
+
"loss": 3.5329,
|
238 |
+
"theoretical_loss": 4.741049741962473,
|
239 |
+
"tokens_seen": 104857600
|
240 |
+
},
|
241 |
+
{
|
242 |
+
"epoch": 0.04,
|
243 |
+
"learning_rate": 0.0004855541895214677,
|
244 |
+
"loss": 3.5329,
|
245 |
+
"theoretical_loss": 4.721042093249051,
|
246 |
+
"tokens_seen": 108134400
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 0.04,
|
250 |
+
"learning_rate": 0.0004849631170796293,
|
251 |
+
"loss": 3.4926,
|
252 |
+
"theoretical_loss": 4.701795828231866,
|
253 |
+
"tokens_seen": 111411200
|
254 |
+
},
|
255 |
+
{
|
256 |
+
"epoch": 0.04,
|
257 |
+
"learning_rate": 0.0004843720446377908,
|
258 |
+
"loss": 3.4835,
|
259 |
+
"theoretical_loss": 4.68326082423593,
|
260 |
+
"tokens_seen": 114688000
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 0.04,
|
264 |
+
"learning_rate": 0.00048378097219595233,
|
265 |
+
"loss": 3.4587,
|
266 |
+
"theoretical_loss": 4.665391572426282,
|
267 |
+
"tokens_seen": 117964800
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 0.04,
|
271 |
+
"learning_rate": 0.00048318989975411385,
|
272 |
+
"loss": 3.4307,
|
273 |
+
"theoretical_loss": 4.648146638719739,
|
274 |
+
"tokens_seen": 121241600
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"epoch": 0.04,
|
278 |
+
"learning_rate": 0.00048259882731227537,
|
279 |
+
"loss": 3.443,
|
280 |
+
"theoretical_loss": 4.631488200339643,
|
281 |
+
"tokens_seen": 124518400
|
282 |
+
},
|
283 |
+
{
|
284 |
+
"epoch": 0.05,
|
285 |
+
"learning_rate": 0.0004820077548704369,
|
286 |
+
"loss": 3.45,
|
287 |
+
"theoretical_loss": 4.615381645715717,
|
288 |
+
"tokens_seen": 127795200
|
289 |
+
},
|
290 |
+
{
|
291 |
+
"epoch": 0.05,
|
292 |
+
"learning_rate": 0.00048141668242859847,
|
293 |
+
"loss": 3.3857,
|
294 |
+
"theoretical_loss": 4.599795227690505,
|
295 |
+
"tokens_seen": 131072000
|
296 |
+
},
|
297 |
+
{
|
298 |
+
"epoch": 0.05,
|
299 |
+
"learning_rate": 0.00048082560998676,
|
300 |
+
"loss": 3.4065,
|
301 |
+
"theoretical_loss": 4.584699761792674,
|
302 |
+
"tokens_seen": 134348800
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 0.05,
|
306 |
+
"learning_rate": 0.0004802345375449215,
|
307 |
+
"loss": 3.3659,
|
308 |
+
"theoretical_loss": 4.570068362778516,
|
309 |
+
"tokens_seen": 137625600
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 0.05,
|
313 |
+
"learning_rate": 0.00047964346510308303,
|
314 |
+
"loss": 3.3781,
|
315 |
+
"theoretical_loss": 4.555876213804037,
|
316 |
+
"tokens_seen": 140902400
|
317 |
+
},
|
318 |
+
{
|
319 |
+
"epoch": 0.05,
|
320 |
+
"learning_rate": 0.00047905239266124455,
|
321 |
+
"loss": 3.3612,
|
322 |
+
"theoretical_loss": 4.542100363530799,
|
323 |
+
"tokens_seen": 144179200
|
324 |
+
},
|
325 |
+
{
|
326 |
+
"epoch": 0.05,
|
327 |
+
"learning_rate": 0.00047846132021940607,
|
328 |
+
"loss": 3.3402,
|
329 |
+
"theoretical_loss": 4.528719547234816,
|
330 |
+
"tokens_seen": 147456000
|
331 |
+
},
|
332 |
+
{
|
333 |
+
"epoch": 0.05,
|
334 |
+
"learning_rate": 0.0004778702477775676,
|
335 |
+
"loss": 3.2899,
|
336 |
+
"theoretical_loss": 4.515714028614996,
|
337 |
+
"tokens_seen": 150732800
|
338 |
+
},
|
339 |
+
{
|
340 |
+
"epoch": 0.06,
|
341 |
+
"learning_rate": 0.00047727917533572917,
|
342 |
+
"loss": 3.3099,
|
343 |
+
"theoretical_loss": 4.503065459513339,
|
344 |
+
"tokens_seen": 154009600
|
345 |
+
},
|
346 |
+
{
|
347 |
+
"epoch": 0.06,
|
348 |
+
"learning_rate": 0.0004766881028938907,
|
349 |
+
"loss": 3.3162,
|
350 |
+
"theoretical_loss": 4.4907567551852665,
|
351 |
+
"tokens_seen": 157286400
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 0.06,
|
355 |
+
"learning_rate": 0.0004760970304520522,
|
356 |
+
"loss": 3.3036,
|
357 |
+
"theoretical_loss": 4.478771983111967,
|
358 |
+
"tokens_seen": 160563200
|
359 |
+
},
|
360 |
+
{
|
361 |
+
"epoch": 0.06,
|
362 |
+
"objective/train/docs_used": 104000,
|
363 |
+
"objective/train/instantaneous_batch_size": 32,
|
364 |
+
"objective/train/instantaneous_microbatch_size": 32768,
|
365 |
+
"objective/train/original_loss": 3.2747654914855957,
|
366 |
+
"objective/train/theoretical_loss": 4.467096263641219,
|
367 |
+
"objective/train/tokens_used": 184300000,
|
368 |
+
"theoretical_loss": 4.467096263641219,
|
369 |
+
"tokens_seen": 163840000
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 0.06,
|
373 |
+
"learning_rate": 0.00047550595801021373,
|
374 |
+
"loss": 3.2837,
|
375 |
+
"theoretical_loss": 4.467096263641219,
|
376 |
+
"tokens_seen": 163840000
|
377 |
+
},
|
378 |
+
{
|
379 |
+
"epoch": 0.06,
|
380 |
+
"learning_rate": 0.00047491488556837525,
|
381 |
+
"loss": 3.2955,
|
382 |
+
"theoretical_loss": 4.455715680989545,
|
383 |
+
"tokens_seen": 167116800
|
384 |
+
},
|
385 |
+
{
|
386 |
+
"epoch": 0.06,
|
387 |
+
"learning_rate": 0.00047432381312653677,
|
388 |
+
"loss": 3.2608,
|
389 |
+
"theoretical_loss": 4.44461720334543,
|
390 |
+
"tokens_seen": 170393600
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"epoch": 0.06,
|
394 |
+
"learning_rate": 0.00047373274068469835,
|
395 |
+
"loss": 3.2026,
|
396 |
+
"theoretical_loss": 4.433788610987646,
|
397 |
+
"tokens_seen": 173670400
|
398 |
+
},
|
399 |
+
{
|
400 |
+
"epoch": 0.06,
|
401 |
+
"learning_rate": 0.00047314166824285987,
|
402 |
+
"loss": 3.2621,
|
403 |
+
"theoretical_loss": 4.42321843148016,
|
404 |
+
"tokens_seen": 176947200
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 0.06,
|
408 |
+
"learning_rate": 0.0004725505958010214,
|
409 |
+
"loss": 3.1999,
|
410 |
+
"theoretical_loss": 4.412895881130142,
|
411 |
+
"tokens_seen": 180224000
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 0.07,
|
415 |
+
"learning_rate": 0.0004719595233591829,
|
416 |
+
"loss": 3.2384,
|
417 |
+
"theoretical_loss": 4.4028108120020795,
|
418 |
+
"tokens_seen": 183500800
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"epoch": 0.07,
|
422 |
+
"learning_rate": 0.00047136845091734443,
|
423 |
+
"loss": 3.244,
|
424 |
+
"theoretical_loss": 4.392953663871862,
|
425 |
+
"tokens_seen": 186777600
|
426 |
+
},
|
427 |
+
{
|
428 |
+
"epoch": 0.07,
|
429 |
+
"learning_rate": 0.00047077737847550595,
|
430 |
+
"loss": 3.2329,
|
431 |
+
"theoretical_loss": 4.383315420582533,
|
432 |
+
"tokens_seen": 190054400
|
433 |
+
},
|
434 |
+
{
|
435 |
+
"epoch": 0.07,
|
436 |
+
"learning_rate": 0.0004701863060336675,
|
437 |
+
"loss": 3.2634,
|
438 |
+
"theoretical_loss": 4.373887570330275,
|
439 |
+
"tokens_seen": 193331200
|
440 |
+
},
|
441 |
+
{
|
442 |
+
"epoch": 0.07,
|
443 |
+
"learning_rate": 0.00046959523359182905,
|
444 |
+
"loss": 3.2143,
|
445 |
+
"theoretical_loss": 4.364662069466704,
|
446 |
+
"tokens_seen": 196608000
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"epoch": 0.07,
|
450 |
+
"learning_rate": 0.00046900416114999057,
|
451 |
+
"loss": 3.2128,
|
452 |
+
"theoretical_loss": 4.355631309453283,
|
453 |
+
"tokens_seen": 199884800
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 0.07,
|
457 |
+
"learning_rate": 0.0004684130887081521,
|
458 |
+
"loss": 3.1675,
|
459 |
+
"theoretical_loss": 4.346788086646671,
|
460 |
+
"tokens_seen": 203161600
|
461 |
+
},
|
462 |
+
{
|
463 |
+
"epoch": 0.07,
|
464 |
+
"learning_rate": 0.0004678220162663136,
|
465 |
+
"loss": 3.1967,
|
466 |
+
"theoretical_loss": 4.33812557463116,
|
467 |
+
"tokens_seen": 206438400
|
468 |
+
},
|
469 |
+
{
|
470 |
+
"epoch": 0.07,
|
471 |
+
"learning_rate": 0.00046723094382447513,
|
472 |
+
"loss": 3.2042,
|
473 |
+
"theoretical_loss": 4.329637298846812,
|
474 |
+
"tokens_seen": 209715200
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 0.08,
|
478 |
+
"learning_rate": 0.00046663987138263665,
|
479 |
+
"loss": 3.1574,
|
480 |
+
"theoretical_loss": 4.321317113290252,
|
481 |
+
"tokens_seen": 212992000
|
482 |
+
},
|
483 |
+
{
|
484 |
+
"epoch": 0.08,
|
485 |
+
"learning_rate": 0.0004660487989407982,
|
486 |
+
"loss": 3.1317,
|
487 |
+
"theoretical_loss": 4.3131591790897925,
|
488 |
+
"tokens_seen": 216268800
|
489 |
+
},
|
490 |
+
{
|
491 |
+
"epoch": 0.08,
|
492 |
+
"learning_rate": 0.00046545772649895975,
|
493 |
+
"loss": 3.1829,
|
494 |
+
"theoretical_loss": 4.305157944778228,
|
495 |
+
"tokens_seen": 219545600
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 0.08,
|
499 |
+
"learning_rate": 0.00046486665405712127,
|
500 |
+
"loss": 3.2073,
|
501 |
+
"theoretical_loss": 4.297308128105687,
|
502 |
+
"tokens_seen": 222822400
|
503 |
+
},
|
504 |
+
{
|
505 |
+
"epoch": 0.08,
|
506 |
+
"learning_rate": 0.0004642755816152828,
|
507 |
+
"loss": 3.1994,
|
508 |
+
"theoretical_loss": 4.2896046992515995,
|
509 |
+
"tokens_seen": 226099200
|
510 |
+
},
|
511 |
+
{
|
512 |
+
"epoch": 0.08,
|
513 |
+
"learning_rate": 0.0004636845091734443,
|
514 |
+
"loss": 3.2312,
|
515 |
+
"theoretical_loss": 4.282042865309616,
|
516 |
+
"tokens_seen": 229376000
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"epoch": 0.08,
|
520 |
+
"learning_rate": 0.00046309343673160583,
|
521 |
+
"loss": 3.2006,
|
522 |
+
"theoretical_loss": 4.274618055932298,
|
523 |
+
"tokens_seen": 232652800
|
524 |
+
},
|
525 |
+
{
|
526 |
+
"epoch": 0.08,
|
527 |
+
"learning_rate": 0.0004625023642897674,
|
528 |
+
"loss": 3.1756,
|
529 |
+
"theoretical_loss": 4.267325910033897,
|
530 |
+
"tokens_seen": 235929600
|
531 |
+
},
|
532 |
+
{
|
533 |
+
"epoch": 0.09,
|
534 |
+
"learning_rate": 0.0004619112918479289,
|
535 |
+
"loss": 3.1158,
|
536 |
+
"theoretical_loss": 4.260162263459744,
|
537 |
+
"tokens_seen": 239206400
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 0.09,
|
541 |
+
"learning_rate": 0.00046132021940609044,
|
542 |
+
"loss": 3.1448,
|
543 |
+
"theoretical_loss": 4.253123137539814,
|
544 |
+
"tokens_seen": 242483200
|
545 |
+
},
|
546 |
+
{
|
547 |
+
"epoch": 0.09,
|
548 |
+
"learning_rate": 0.00046072914696425197,
|
549 |
+
"loss": 3.166,
|
550 |
+
"theoretical_loss": 4.246204728452055,
|
551 |
+
"tokens_seen": 245760000
|
552 |
+
},
|
553 |
+
{
|
554 |
+
"epoch": 0.09,
|
555 |
+
"learning_rate": 0.0004601380745224135,
|
556 |
+
"loss": 3.1762,
|
557 |
+
"theoretical_loss": 4.239403397328261,
|
558 |
+
"tokens_seen": 249036800
|
559 |
+
},
|
560 |
+
{
|
561 |
+
"epoch": 0.09,
|
562 |
+
"learning_rate": 0.000459547002080575,
|
563 |
+
"loss": 3.1442,
|
564 |
+
"theoretical_loss": 4.232715661041632,
|
565 |
+
"tokens_seen": 252313600
|
566 |
+
},
|
567 |
+
{
|
568 |
+
"epoch": 0.09,
|
569 |
+
"learning_rate": 0.0004589559296387366,
|
570 |
+
"loss": 3.1504,
|
571 |
+
"theoretical_loss": 4.226138183620867,
|
572 |
+
"tokens_seen": 255590400
|
573 |
+
},
|
574 |
+
{
|
575 |
+
"epoch": 0.09,
|
576 |
+
"learning_rate": 0.0004583648571968981,
|
577 |
+
"loss": 3.1099,
|
578 |
+
"theoretical_loss": 4.219667768240775,
|
579 |
+
"tokens_seen": 258867200
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 0.09,
|
583 |
+
"learning_rate": 0.0004577737847550596,
|
584 |
+
"loss": 3.114,
|
585 |
+
"theoretical_loss": 4.213301349743924,
|
586 |
+
"tokens_seen": 262144000
|
587 |
+
},
|
588 |
+
{
|
589 |
+
"epoch": 0.09,
|
590 |
+
"learning_rate": 0.0004571827123132211,
|
591 |
+
"loss": 3.0703,
|
592 |
+
"theoretical_loss": 4.20703598765197,
|
593 |
+
"tokens_seen": 265420800
|
594 |
+
},
|
595 |
+
{
|
596 |
+
"epoch": 0.1,
|
597 |
+
"learning_rate": 0.0004565916398713826,
|
598 |
+
"loss": 3.0383,
|
599 |
+
"theoretical_loss": 4.2008688596290025,
|
600 |
+
"tokens_seen": 268697600
|
601 |
+
},
|
602 |
+
{
|
603 |
+
"epoch": 0.1,
|
604 |
+
"learning_rate": 0.00045600056742954413,
|
605 |
+
"loss": 2.9881,
|
606 |
+
"theoretical_loss": 4.194797255362549,
|
607 |
+
"tokens_seen": 271974400
|
608 |
+
},
|
609 |
+
{
|
610 |
+
"epoch": 0.1,
|
611 |
+
"learning_rate": 0.0004554094949877057,
|
612 |
+
"loss": 3.068,
|
613 |
+
"theoretical_loss": 4.188818570830883,
|
614 |
+
"tokens_seen": 275251200
|
615 |
+
},
|
616 |
+
{
|
617 |
+
"epoch": 0.1,
|
618 |
+
"learning_rate": 0.00045481842254586723,
|
619 |
+
"loss": 3.0823,
|
620 |
+
"theoretical_loss": 4.182930302927963,
|
621 |
+
"tokens_seen": 278528000
|
622 |
+
},
|
623 |
+
{
|
624 |
+
"epoch": 0.1,
|
625 |
+
"learning_rate": 0.00045422735010402875,
|
626 |
+
"loss": 3.0393,
|
627 |
+
"theoretical_loss": 4.17713004441978,
|
628 |
+
"tokens_seen": 281804800
|
629 |
+
},
|
630 |
+
{
|
631 |
+
"epoch": 0.1,
|
632 |
+
"learning_rate": 0.00045363627766219027,
|
633 |
+
"loss": 3.0775,
|
634 |
+
"theoretical_loss": 4.1714154792080915,
|
635 |
+
"tokens_seen": 285081600
|
636 |
+
},
|
637 |
+
{
|
638 |
+
"epoch": 0.1,
|
639 |
+
"learning_rate": 0.0004530452052203518,
|
640 |
+
"loss": 3.0876,
|
641 |
+
"theoretical_loss": 4.165784377879517,
|
642 |
+
"tokens_seen": 288358400
|
643 |
+
},
|
644 |
+
{
|
645 |
+
"epoch": 0.1,
|
646 |
+
"learning_rate": 0.0004524541327785133,
|
647 |
+
"loss": 3.0473,
|
648 |
+
"theoretical_loss": 4.160234593519768,
|
649 |
+
"tokens_seen": 291635200
|
650 |
+
},
|
651 |
+
{
|
652 |
+
"epoch": 0.11,
|
653 |
+
"learning_rate": 0.00045186306033667483,
|
654 |
+
"loss": 3.0706,
|
655 |
+
"theoretical_loss": 4.15476405777444,
|
656 |
+
"tokens_seen": 294912000
|
657 |
+
},
|
658 |
+
{
|
659 |
+
"epoch": 0.11,
|
660 |
+
"learning_rate": 0.0004512719878948364,
|
661 |
+
"loss": 3.0424,
|
662 |
+
"theoretical_loss": 4.149370777139286,
|
663 |
+
"tokens_seen": 298188800
|
664 |
+
},
|
665 |
+
{
|
666 |
+
"epoch": 0.11,
|
667 |
+
"learning_rate": 0.00045068091545299793,
|
668 |
+
"loss": 3.04,
|
669 |
+
"theoretical_loss": 4.144052829464249,
|
670 |
+
"tokens_seen": 301465600
|
671 |
+
},
|
672 |
+
{
|
673 |
+
"epoch": 0.11,
|
674 |
+
"learning_rate": 0.00045008984301115945,
|
675 |
+
"loss": 3.0118,
|
676 |
+
"theoretical_loss": 4.138808360656742,
|
677 |
+
"tokens_seen": 304742400
|
678 |
+
},
|
679 |
+
{
|
680 |
+
"epoch": 0.11,
|
681 |
+
"learning_rate": 0.00044949877056932097,
|
682 |
+
"loss": 3.0093,
|
683 |
+
"theoretical_loss": 4.133635581570836,
|
684 |
+
"tokens_seen": 308019200
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"epoch": 0.11,
|
688 |
+
"learning_rate": 0.0004489076981274825,
|
689 |
+
"loss": 2.9812,
|
690 |
+
"theoretical_loss": 4.128532765070004,
|
691 |
+
"tokens_seen": 311296000
|
692 |
+
},
|
693 |
+
{
|
694 |
+
"epoch": 0.11,
|
695 |
+
"learning_rate": 0.000448316625685644,
|
696 |
+
"loss": 3.0051,
|
697 |
+
"theoretical_loss": 4.123498243252032,
|
698 |
+
"tokens_seen": 314572800
|
699 |
+
},
|
700 |
+
{
|
701 |
+
"epoch": 0.11,
|
702 |
+
"learning_rate": 0.0004477255532438056,
|
703 |
+
"loss": 2.9872,
|
704 |
+
"theoretical_loss": 4.118530404825556,
|
705 |
+
"tokens_seen": 317849600
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"epoch": 0.11,
|
709 |
+
"learning_rate": 0.0004471344808019671,
|
710 |
+
"loss": 3.0065,
|
711 |
+
"theoretical_loss": 4.113627692628464,
|
712 |
+
"tokens_seen": 321126400
|
713 |
+
},
|
714 |
+
{
|
715 |
+
"epoch": 0.12,
|
716 |
+
"learning_rate": 0.00044654340836012863,
|
717 |
+
"loss": 3.0122,
|
718 |
+
"theoretical_loss": 4.108788601279149,
|
719 |
+
"tokens_seen": 324403200
|
720 |
+
},
|
721 |
+
{
|
722 |
+
"debugging/Self-BLEU-5": 0.5365128506817183,
|
723 |
+
"debugging/distinct-1-grams": 0.7612814402327299,
|
724 |
+
"debugging/distinct-2-grams": 0.9694583753853511,
|
725 |
+
"debugging/entropy-1-grams": 6.003629944255698,
|
726 |
+
"debugging/entropy-2-grams": 7.054987089269872,
|
727 |
+
"debugging/length": 495.25,
|
728 |
+
"debugging/num_segments": 16,
|
729 |
+
"epoch": 0.12,
|
730 |
+
"objective/train/docs_used": 197327,
|
731 |
+
"objective/train/instantaneous_batch_size": 32,
|
732 |
+
"objective/train/instantaneous_microbatch_size": 32768,
|
733 |
+
"objective/train/original_loss": 2.9225306510925293,
|
734 |
+
"objective/train/theoretical_loss": 4.10401167495222,
|
735 |
+
"objective/train/tokens_used": 348140000,
|
736 |
+
"theoretical_loss": 4.10401167495222,
|
737 |
+
"tokens_seen": 327680000
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 0.12,
|
741 |
+
"learning_rate": 0.00044595233591829015,
|
742 |
+
"loss": 3.0423,
|
743 |
+
"theoretical_loss": 4.10401167495222,
|
744 |
+
"tokens_seen": 327680000
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 0.12,
|
748 |
+
"learning_rate": 0.00044536126347645167,
|
749 |
+
"loss": 3.041,
|
750 |
+
"theoretical_loss": 4.099295505270921,
|
751 |
+
"tokens_seen": 330956800
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 0.12,
|
755 |
+
"learning_rate": 0.0004447701910346132,
|
756 |
+
"loss": 2.9867,
|
757 |
+
"theoretical_loss": 4.094638729309031,
|
758 |
+
"tokens_seen": 334233600
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 0.12,
|
762 |
+
"learning_rate": 0.00044417911859277476,
|
763 |
+
"loss": 2.9992,
|
764 |
+
"theoretical_loss": 4.090040027695556,
|
765 |
+
"tokens_seen": 337510400
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 0.12,
|
769 |
+
"learning_rate": 0.0004435880461509363,
|
770 |
+
"loss": 2.9584,
|
771 |
+
"theoretical_loss": 4.085498122815992,
|
772 |
+
"tokens_seen": 340787200
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 0.12,
|
776 |
+
"learning_rate": 0.0004429969737090978,
|
777 |
+
"loss": 2.9834,
|
778 |
+
"theoretical_loss": 4.081011777104333,
|
779 |
+
"tokens_seen": 344064000
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 0.12,
|
783 |
+
"learning_rate": 0.0004424059012672593,
|
784 |
+
"loss": 2.9805,
|
785 |
+
"theoretical_loss": 4.076579791420469,
|
786 |
+
"tokens_seen": 347340800
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 0.13,
|
790 |
+
"learning_rate": 0.00044181482882542085,
|
791 |
+
"loss": 2.9998,
|
792 |
+
"theoretical_loss": 4.0722010035079155,
|
793 |
+
"tokens_seen": 350617600
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 0.13,
|
797 |
+
"learning_rate": 0.00044122375638358237,
|
798 |
+
"loss": 2.9906,
|
799 |
+
"theoretical_loss": 4.067874286527197,
|
800 |
+
"tokens_seen": 353894400
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 0.13,
|
804 |
+
"learning_rate": 0.0004406326839417439,
|
805 |
+
"loss": 2.9923,
|
806 |
+
"theoretical_loss": 4.063598547660519,
|
807 |
+
"tokens_seen": 357171200
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 0.13,
|
811 |
+
"learning_rate": 0.00044004161149990546,
|
812 |
+
"loss": 3.0199,
|
813 |
+
"theoretical_loss": 4.05937272678363,
|
814 |
+
"tokens_seen": 360448000
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 0.13,
|
818 |
+
"learning_rate": 0.000439450539058067,
|
819 |
+
"loss": 3.0079,
|
820 |
+
"theoretical_loss": 4.055195795201069,
|
821 |
+
"tokens_seen": 363724800
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 0.13,
|
825 |
+
"learning_rate": 0.0004388594666162285,
|
826 |
+
"loss": 2.9757,
|
827 |
+
"theoretical_loss": 4.051066754441235,
|
828 |
+
"tokens_seen": 367001600
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 0.13,
|
832 |
+
"learning_rate": 0.00043826839417439,
|
833 |
+
"loss": 2.9698,
|
834 |
+
"theoretical_loss": 4.04698463510794,
|
835 |
+
"tokens_seen": 370278400
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 0.13,
|
839 |
+
"learning_rate": 0.00043767732173255155,
|
840 |
+
"loss": 2.9735,
|
841 |
+
"theoretical_loss": 4.042948495785312,
|
842 |
+
"tokens_seen": 373555200
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 0.13,
|
846 |
+
"learning_rate": 0.00043708624929071307,
|
847 |
+
"loss": 2.9327,
|
848 |
+
"theoretical_loss": 4.038957421993153,
|
849 |
+
"tokens_seen": 376832000
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.14,
|
853 |
+
"learning_rate": 0.00043649517684887464,
|
854 |
+
"loss": 2.9425,
|
855 |
+
"theoretical_loss": 4.035010525189982,
|
856 |
+
"tokens_seen": 380108800
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 0.14,
|
860 |
+
"learning_rate": 0.00043590410440703616,
|
861 |
+
"loss": 2.9746,
|
862 |
+
"theoretical_loss": 4.031106941821218,
|
863 |
+
"tokens_seen": 383385600
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 0.14,
|
867 |
+
"learning_rate": 0.0004353130319651977,
|
868 |
+
"loss": 2.9371,
|
869 |
+
"theoretical_loss": 4.027245832410079,
|
870 |
+
"tokens_seen": 386662400
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 0.14,
|
874 |
+
"learning_rate": 0.0004347219595233592,
|
875 |
+
"loss": 2.9136,
|
876 |
+
"theoretical_loss": 4.023426380688943,
|
877 |
+
"tokens_seen": 389939200
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 0.14,
|
881 |
+
"learning_rate": 0.0004341308870815207,
|
882 |
+
"loss": 2.9586,
|
883 |
+
"theoretical_loss": 4.019647792769048,
|
884 |
+
"tokens_seen": 393216000
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 0.14,
|
888 |
+
"learning_rate": 0.00043353981463968225,
|
889 |
+
"loss": 3.0316,
|
890 |
+
"theoretical_loss": 4.015909296346521,
|
891 |
+
"tokens_seen": 396492800
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 0.14,
|
895 |
+
"learning_rate": 0.0004329487421978438,
|
896 |
+
"loss": 2.974,
|
897 |
+
"theoretical_loss": 4.012210139942894,
|
898 |
+
"tokens_seen": 399769600
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 0.14,
|
902 |
+
"learning_rate": 0.00043235766975600534,
|
903 |
+
"loss": 3.0154,
|
904 |
+
"theoretical_loss": 4.008549592178291,
|
905 |
+
"tokens_seen": 403046400
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 0.15,
|
909 |
+
"learning_rate": 0.00043176659731416686,
|
910 |
+
"loss": 2.9866,
|
911 |
+
"theoretical_loss": 4.004926941075674,
|
912 |
+
"tokens_seen": 406323200
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 0.15,
|
916 |
+
"learning_rate": 0.0004311755248723284,
|
917 |
+
"loss": 3.0274,
|
918 |
+
"theoretical_loss": 4.001341493394558,
|
919 |
+
"tokens_seen": 409600000
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 0.15,
|
923 |
+
"learning_rate": 0.0004305844524304899,
|
924 |
+
"loss": 3.0314,
|
925 |
+
"theoretical_loss": 3.997792573992726,
|
926 |
+
"tokens_seen": 412876800
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 0.15,
|
930 |
+
"learning_rate": 0.0004299933799886514,
|
931 |
+
"loss": 2.9786,
|
932 |
+
"theoretical_loss": 3.994279525214554,
|
933 |
+
"tokens_seen": 416153600
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 0.15,
|
937 |
+
"learning_rate": 0.00042940230754681295,
|
938 |
+
"loss": 3.0252,
|
939 |
+
"theoretical_loss": 3.990801706304647,
|
940 |
+
"tokens_seen": 419430400
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 0.15,
|
944 |
+
"learning_rate": 0.0004288112351049745,
|
945 |
+
"loss": 2.9866,
|
946 |
+
"theoretical_loss": 3.987358492845532,
|
947 |
+
"tokens_seen": 422707200
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 0.15,
|
951 |
+
"learning_rate": 0.000428220162663136,
|
952 |
+
"loss": 3.0081,
|
953 |
+
"theoretical_loss": 3.9839492762182647,
|
954 |
+
"tokens_seen": 425984000
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 0.15,
|
958 |
+
"learning_rate": 0.0004276290902212975,
|
959 |
+
"loss": 2.9837,
|
960 |
+
"theoretical_loss": 3.9805734630848306,
|
961 |
+
"tokens_seen": 429260800
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 0.15,
|
965 |
+
"learning_rate": 0.00042703801777945903,
|
966 |
+
"loss": 2.9656,
|
967 |
+
"theoretical_loss": 3.9772304748913054,
|
968 |
+
"tokens_seen": 432537600
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 0.16,
|
972 |
+
"learning_rate": 0.00042644694533762055,
|
973 |
+
"loss": 2.9527,
|
974 |
+
"theoretical_loss": 3.973919747390801,
|
975 |
+
"tokens_seen": 435814400
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 0.16,
|
979 |
+
"learning_rate": 0.00042585587289578207,
|
980 |
+
"loss": 2.9704,
|
981 |
+
"theoretical_loss": 3.9706407301852487,
|
982 |
+
"tokens_seen": 439091200
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 0.16,
|
986 |
+
"learning_rate": 0.00042526480045394365,
|
987 |
+
"loss": 2.9481,
|
988 |
+
"theoretical_loss": 3.9673928862851655,
|
989 |
+
"tokens_seen": 442368000
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 0.16,
|
993 |
+
"learning_rate": 0.00042467372801210517,
|
994 |
+
"loss": 2.9408,
|
995 |
+
"theoretical_loss": 3.9641756916865463,
|
996 |
+
"tokens_seen": 445644800
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 0.16,
|
1000 |
+
"learning_rate": 0.0004240826555702667,
|
1001 |
+
"loss": 2.9444,
|
1002 |
+
"theoretical_loss": 3.960988634964113,
|
1003 |
+
"tokens_seen": 448921600
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 0.16,
|
1007 |
+
"learning_rate": 0.0004234915831284282,
|
1008 |
+
"loss": 2.9224,
|
1009 |
+
"theoretical_loss": 3.9578312168801597,
|
1010 |
+
"tokens_seen": 452198400
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 0.16,
|
1014 |
+
"learning_rate": 0.00042290051068658973,
|
1015 |
+
"loss": 2.9187,
|
1016 |
+
"theoretical_loss": 3.954702950008308,
|
1017 |
+
"tokens_seen": 455475200
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 0.16,
|
1021 |
+
"learning_rate": 0.00042230943824475125,
|
1022 |
+
"loss": 2.9027,
|
1023 |
+
"theoretical_loss": 3.9516033583714734,
|
1024 |
+
"tokens_seen": 458752000
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 0.17,
|
1028 |
+
"learning_rate": 0.0004217183658029128,
|
1029 |
+
"loss": 2.9277,
|
1030 |
+
"theoretical_loss": 3.9485319770934355,
|
1031 |
+
"tokens_seen": 462028800
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 0.17,
|
1035 |
+
"learning_rate": 0.00042112729336107435,
|
1036 |
+
"loss": 2.9109,
|
1037 |
+
"theoretical_loss": 3.945488352063391,
|
1038 |
+
"tokens_seen": 465305600
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 0.17,
|
1042 |
+
"learning_rate": 0.00042053622091923587,
|
1043 |
+
"loss": 2.8861,
|
1044 |
+
"theoretical_loss": 3.942472039612926,
|
1045 |
+
"tokens_seen": 468582400
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 0.17,
|
1049 |
+
"learning_rate": 0.0004199451484773974,
|
1050 |
+
"loss": 2.8901,
|
1051 |
+
"theoretical_loss": 3.939482606204863,
|
1052 |
+
"tokens_seen": 471859200
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 0.17,
|
1056 |
+
"learning_rate": 0.0004193540760355589,
|
1057 |
+
"loss": 2.9065,
|
1058 |
+
"theoretical_loss": 3.936519628133466,
|
1059 |
+
"tokens_seen": 475136000
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 0.17,
|
1063 |
+
"learning_rate": 0.00041876300359372043,
|
1064 |
+
"loss": 2.9422,
|
1065 |
+
"theoretical_loss": 3.9335826912355114,
|
1066 |
+
"tokens_seen": 478412800
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 0.17,
|
1070 |
+
"learning_rate": 0.00041817193115188195,
|
1071 |
+
"loss": 2.9805,
|
1072 |
+
"theoretical_loss": 3.93067139061177,
|
1073 |
+
"tokens_seen": 481689600
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 0.17,
|
1077 |
+
"learning_rate": 0.0004175808587100435,
|
1078 |
+
"loss": 2.9047,
|
1079 |
+
"theoretical_loss": 3.927785330358441,
|
1080 |
+
"tokens_seen": 484966400
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 0.17,
|
1084 |
+
"learning_rate": 0.00041698978626820505,
|
1085 |
+
"loss": 2.8916,
|
1086 |
+
"theoretical_loss": 3.9249241233081333,
|
1087 |
+
"tokens_seen": 488243200
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 0.18,
|
1091 |
+
"objective/train/docs_used": 287192,
|
1092 |
+
"objective/train/instantaneous_batch_size": 32,
|
1093 |
+
"objective/train/instantaneous_microbatch_size": 32768,
|
1094 |
+
"objective/train/original_loss": 3.0162453651428223,
|
1095 |
+
"objective/train/theoretical_loss": 3.92208739077998,
|
1096 |
+
"objective/train/tokens_used": 511980000,
|
1097 |
+
"theoretical_loss": 3.92208739077998,
|
1098 |
+
"tokens_seen": 491520000
|
1099 |
+
},
|
1100 |
+
{
|
1101 |
+
"epoch": 0.18,
|
1102 |
+
"learning_rate": 0.00041639871382636657,
|
1103 |
+
"loss": 2.8641,
|
1104 |
+
"theoretical_loss": 3.92208739077998,
|
1105 |
+
"tokens_seen": 491520000
|
1106 |
+
},
|
1107 |
+
{
|
1108 |
+
"epoch": 0.18,
|
1109 |
+
"learning_rate": 0.0004158076413845281,
|
1110 |
+
"loss": 2.8749,
|
1111 |
+
"theoretical_loss": 3.919274762338519,
|
1112 |
+
"tokens_seen": 494796800
|
1113 |
+
},
|
1114 |
+
{
|
1115 |
+
"epoch": 0.18,
|
1116 |
+
"learning_rate": 0.0004152165689426896,
|
1117 |
+
"loss": 2.8998,
|
1118 |
+
"theoretical_loss": 3.9164858755609613,
|
1119 |
+
"tokens_seen": 498073600
|
1120 |
+
},
|
1121 |
+
{
|
1122 |
+
"epoch": 0.18,
|
1123 |
+
"learning_rate": 0.00041462549650085113,
|
1124 |
+
"loss": 2.8889,
|
1125 |
+
"theoretical_loss": 3.9137203758125176,
|
1126 |
+
"tokens_seen": 501350400
|
1127 |
+
},
|
1128 |
+
{
|
1129 |
+
"epoch": 0.18,
|
1130 |
+
"learning_rate": 0.0004140344240590127,
|
1131 |
+
"loss": 2.8474,
|
1132 |
+
"theoretical_loss": 3.910977916029439,
|
1133 |
+
"tokens_seen": 504627200
|
1134 |
+
},
|
1135 |
+
{
|
1136 |
+
"epoch": 0.18,
|
1137 |
+
"learning_rate": 0.0004134433516171742,
|
1138 |
+
"loss": 2.8623,
|
1139 |
+
"theoretical_loss": 3.908258156509472,
|
1140 |
+
"tokens_seen": 507904000
|
1141 |
+
},
|
1142 |
+
{
|
1143 |
+
"epoch": 0.18,
|
1144 |
+
"learning_rate": 0.00041285227917533575,
|
1145 |
+
"loss": 2.85,
|
1146 |
+
"theoretical_loss": 3.905560764709417,
|
1147 |
+
"tokens_seen": 511180800
|
1148 |
+
},
|
1149 |
+
{
|
1150 |
+
"epoch": 0.18,
|
1151 |
+
"learning_rate": 0.00041226120673349727,
|
1152 |
+
"loss": 2.8285,
|
1153 |
+
"theoretical_loss": 3.9028854150495143,
|
1154 |
+
"tokens_seen": 514457600
|
1155 |
+
},
|
1156 |
+
{
|
1157 |
+
"epoch": 0.18,
|
1158 |
+
"learning_rate": 0.0004116701342916588,
|
1159 |
+
"loss": 2.8533,
|
1160 |
+
"theoretical_loss": 3.9002317887243834,
|
1161 |
+
"tokens_seen": 517734400
|
1162 |
+
},
|
1163 |
+
{
|
1164 |
+
"epoch": 0.19,
|
1165 |
+
"learning_rate": 0.0004110790618498203,
|
1166 |
+
"loss": 2.8568,
|
1167 |
+
"theoretical_loss": 3.897599573520247,
|
1168 |
+
"tokens_seen": 521011200
|
1169 |
+
},
|
1170 |
+
{
|
1171 |
+
"epoch": 0.19,
|
1172 |
+
"learning_rate": 0.0004104879894079819,
|
1173 |
+
"loss": 2.8888,
|
1174 |
+
"theoretical_loss": 3.8949884636382106,
|
1175 |
+
"tokens_seen": 524288000
|
1176 |
+
},
|
1177 |
+
{
|
1178 |
+
"epoch": 0.19,
|
1179 |
+
"learning_rate": 0.0004098969169661434,
|
1180 |
+
"loss": 2.8691,
|
1181 |
+
"theoretical_loss": 3.892398159523345,
|
1182 |
+
"tokens_seen": 527564800
|
1183 |
+
},
|
1184 |
+
{
|
1185 |
+
"epoch": 0.19,
|
1186 |
+
"learning_rate": 0.0004093058445243049,
|
1187 |
+
"loss": 2.9266,
|
1188 |
+
"theoretical_loss": 3.889828367699349,
|
1189 |
+
"tokens_seen": 530841600
|
1190 |
+
},
|
1191 |
+
{
|
1192 |
+
"epoch": 0.19,
|
1193 |
+
"learning_rate": 0.00040871477208246645,
|
1194 |
+
"loss": 2.8666,
|
1195 |
+
"theoretical_loss": 3.8872788006085894,
|
1196 |
+
"tokens_seen": 534118400
|
1197 |
+
},
|
1198 |
+
{
|
1199 |
+
"epoch": 0.19,
|
1200 |
+
"learning_rate": 0.00040812369964062797,
|
1201 |
+
"loss": 2.8944,
|
1202 |
+
"theoretical_loss": 3.8847491764572926,
|
1203 |
+
"tokens_seen": 537395200
|
1204 |
+
},
|
1205 |
+
{
|
1206 |
+
"epoch": 0.19,
|
1207 |
+
"learning_rate": 0.0004075326271987895,
|
1208 |
+
"loss": 2.8746,
|
1209 |
+
"theoretical_loss": 3.882239219065708,
|
1210 |
+
"tokens_seen": 540672000
|
1211 |
+
},
|
1212 |
+
{
|
1213 |
+
"epoch": 0.19,
|
1214 |
+
"learning_rate": 0.000406941554756951,
|
1215 |
+
"loss": 2.9009,
|
1216 |
+
"theoretical_loss": 3.879748657723039,
|
1217 |
+
"tokens_seen": 543948800
|
1218 |
+
},
|
1219 |
+
{
|
1220 |
+
"epoch": 0.2,
|
1221 |
+
"learning_rate": 0.0004063504823151126,
|
1222 |
+
"loss": 2.9279,
|
1223 |
+
"theoretical_loss": 3.8772772270469824,
|
1224 |
+
"tokens_seen": 547225600
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 0.2,
|
1228 |
+
"learning_rate": 0.0004057594098732741,
|
1229 |
+
"loss": 2.8889,
|
1230 |
+
"theoretical_loss": 3.8748246668476827,
|
1231 |
+
"tokens_seen": 550502400
|
1232 |
+
},
|
1233 |
+
{
|
1234 |
+
"epoch": 0.2,
|
1235 |
+
"learning_rate": 0.0004051683374314356,
|
1236 |
+
"loss": 2.9038,
|
1237 |
+
"theoretical_loss": 3.8723907219959486,
|
1238 |
+
"tokens_seen": 553779200
|
1239 |
+
},
|
1240 |
+
{
|
1241 |
+
"epoch": 0.2,
|
1242 |
+
"learning_rate": 0.00040457726498959715,
|
1243 |
+
"loss": 2.9203,
|
1244 |
+
"theoretical_loss": 3.869975142295573,
|
1245 |
+
"tokens_seen": 557056000
|
1246 |
+
},
|
1247 |
+
{
|
1248 |
+
"epoch": 0.2,
|
1249 |
+
"learning_rate": 0.00040398619254775867,
|
1250 |
+
"loss": 2.9695,
|
1251 |
+
"theoretical_loss": 3.8675776823595998,
|
1252 |
+
"tokens_seen": 560332800
|
1253 |
+
},
|
1254 |
+
{
|
1255 |
+
"epoch": 0.2,
|
1256 |
+
"learning_rate": 0.00040340694155475694,
|
1257 |
+
"loss": 2.9455,
|
1258 |
+
"theoretical_loss": 3.8651981014904027,
|
1259 |
+
"tokens_seen": 563609600
|
1260 |
+
},
|
1261 |
+
{
|
1262 |
+
"epoch": 0.2,
|
1263 |
+
"learning_rate": 0.0004028158691129185,
|
1264 |
+
"loss": 2.9359,
|
1265 |
+
"theoretical_loss": 3.8628361635634265,
|
1266 |
+
"tokens_seen": 566886400
|
1267 |
+
},
|
1268 |
+
{
|
1269 |
+
"epoch": 0.2,
|
1270 |
+
"learning_rate": 0.00040222479667108003,
|
1271 |
+
"loss": 2.9886,
|
1272 |
+
"theoretical_loss": 3.8604916369144666,
|
1273 |
+
"tokens_seen": 570163200
|
1274 |
+
},
|
1275 |
+
{
|
1276 |
+
"epoch": 0.2,
|
1277 |
+
"learning_rate": 0.00040163372422924155,
|
1278 |
+
"loss": 2.9385,
|
1279 |
+
"theoretical_loss": 3.858164294230354,
|
1280 |
+
"tokens_seen": 573440000
|
1281 |
+
},
|
1282 |
+
{
|
1283 |
+
"epoch": 0.21,
|
1284 |
+
"learning_rate": 0.0004010426517874031,
|
1285 |
+
"loss": 2.9009,
|
1286 |
+
"theoretical_loss": 3.85585391244293,
|
1287 |
+
"tokens_seen": 576716800
|
1288 |
+
},
|
1289 |
+
{
|
1290 |
+
"epoch": 0.21,
|
1291 |
+
"learning_rate": 0.0004004515793455646,
|
1292 |
+
"loss": 2.8993,
|
1293 |
+
"theoretical_loss": 3.8535602726261864,
|
1294 |
+
"tokens_seen": 579993600
|
1295 |
+
},
|
1296 |
+
{
|
1297 |
+
"epoch": 0.21,
|
1298 |
+
"learning_rate": 0.0003998605069037261,
|
1299 |
+
"loss": 2.9056,
|
1300 |
+
"theoretical_loss": 3.851283159896468,
|
1301 |
+
"tokens_seen": 583270400
|
1302 |
+
},
|
1303 |
+
{
|
1304 |
+
"epoch": 0.21,
|
1305 |
+
"learning_rate": 0.0003992694344618877,
|
1306 |
+
"loss": 2.8781,
|
1307 |
+
"theoretical_loss": 3.8490223633156173,
|
1308 |
+
"tokens_seen": 586547200
|
1309 |
+
},
|
1310 |
+
{
|
1311 |
+
"epoch": 0.21,
|
1312 |
+
"learning_rate": 0.0003986783620200492,
|
1313 |
+
"loss": 2.8614,
|
1314 |
+
"theoretical_loss": 3.846777675796974,
|
1315 |
+
"tokens_seen": 589824000
|
1316 |
+
},
|
1317 |
+
{
|
1318 |
+
"epoch": 0.21,
|
1319 |
+
"learning_rate": 0.00039808728957821073,
|
1320 |
+
"loss": 2.8569,
|
1321 |
+
"theoretical_loss": 3.844548894014116,
|
1322 |
+
"tokens_seen": 593100800
|
1323 |
+
},
|
1324 |
+
{
|
1325 |
+
"epoch": 0.21,
|
1326 |
+
"learning_rate": 0.00039749621713637225,
|
1327 |
+
"loss": 2.8681,
|
1328 |
+
"theoretical_loss": 3.8423358183122582,
|
1329 |
+
"tokens_seen": 596377600
|
1330 |
+
},
|
1331 |
+
{
|
1332 |
+
"epoch": 0.21,
|
1333 |
+
"learning_rate": 0.0003969051446945338,
|
1334 |
+
"loss": 2.8386,
|
1335 |
+
"theoretical_loss": 3.840138252622208,
|
1336 |
+
"tokens_seen": 599654400
|
1337 |
+
},
|
1338 |
+
{
|
1339 |
+
"epoch": 0.22,
|
1340 |
+
"learning_rate": 0.0003963140722526953,
|
1341 |
+
"loss": 2.877,
|
1342 |
+
"theoretical_loss": 3.837956004376799,
|
1343 |
+
"tokens_seen": 602931200
|
1344 |
+
},
|
1345 |
+
{
|
1346 |
+
"epoch": 0.22,
|
1347 |
+
"learning_rate": 0.00039572299981085687,
|
1348 |
+
"loss": 2.8372,
|
1349 |
+
"theoretical_loss": 3.8357888844297094,
|
1350 |
+
"tokens_seen": 606208000
|
1351 |
+
},
|
1352 |
+
{
|
1353 |
+
"epoch": 0.22,
|
1354 |
+
"learning_rate": 0.0003951319273690184,
|
1355 |
+
"loss": 2.8638,
|
1356 |
+
"theoretical_loss": 3.8336367069765958,
|
1357 |
+
"tokens_seen": 609484800
|
1358 |
+
},
|
1359 |
+
{
|
1360 |
+
"epoch": 0.22,
|
1361 |
+
"learning_rate": 0.0003945408549271799,
|
1362 |
+
"loss": 2.8071,
|
1363 |
+
"theoretical_loss": 3.8314992894784536,
|
1364 |
+
"tokens_seen": 612761600
|
1365 |
+
},
|
1366 |
+
{
|
1367 |
+
"epoch": 0.22,
|
1368 |
+
"learning_rate": 0.00039394978248534143,
|
1369 |
+
"loss": 2.8512,
|
1370 |
+
"theoretical_loss": 3.829376452587134,
|
1371 |
+
"tokens_seen": 616038400
|
1372 |
+
},
|
1373 |
+
{
|
1374 |
+
"epoch": 0.22,
|
1375 |
+
"learning_rate": 0.00039335871004350295,
|
1376 |
+
"loss": 2.8688,
|
1377 |
+
"theoretical_loss": 3.827268020072948,
|
1378 |
+
"tokens_seen": 619315200
|
1379 |
+
},
|
1380 |
+
{
|
1381 |
+
"epoch": 0.22,
|
1382 |
+
"learning_rate": 0.0003927676376016645,
|
1383 |
+
"loss": 2.8993,
|
1384 |
+
"theoretical_loss": 3.8251738187542843,
|
1385 |
+
"tokens_seen": 622592000
|
1386 |
+
},
|
1387 |
+
{
|
1388 |
+
"epoch": 0.22,
|
1389 |
+
"learning_rate": 0.000392176565159826,
|
1390 |
+
"loss": 2.8856,
|
1391 |
+
"theoretical_loss": 3.8230936784291787,
|
1392 |
+
"tokens_seen": 625868800
|
1393 |
+
},
|
1394 |
+
{
|
1395 |
+
"epoch": 0.22,
|
1396 |
+
"learning_rate": 0.00039158549271798757,
|
1397 |
+
"loss": 2.918,
|
1398 |
+
"theoretical_loss": 3.8210274318087656,
|
1399 |
+
"tokens_seen": 629145600
|
1400 |
+
},
|
1401 |
+
{
|
1402 |
+
"epoch": 0.23,
|
1403 |
+
"learning_rate": 0.0003909944202761491,
|
1404 |
+
"loss": 2.8609,
|
1405 |
+
"theoretical_loss": 3.818974914452557,
|
1406 |
+
"tokens_seen": 632422400
|
1407 |
+
},
|
1408 |
+
{
|
1409 |
+
"epoch": 0.23,
|
1410 |
+
"learning_rate": 0.00039040334783431056,
|
1411 |
+
"loss": 2.8562,
|
1412 |
+
"theoretical_loss": 3.8169359647054835,
|
1413 |
+
"tokens_seen": 635699200
|
1414 |
+
},
|
1415 |
+
{
|
1416 |
+
"epoch": 0.23,
|
1417 |
+
"learning_rate": 0.0003898122753924721,
|
1418 |
+
"loss": 2.8437,
|
1419 |
+
"theoretical_loss": 3.8149104236366433,
|
1420 |
+
"tokens_seen": 638976000
|
1421 |
+
},
|
1422 |
+
{
|
1423 |
+
"epoch": 0.23,
|
1424 |
+
"learning_rate": 0.0003892212029506336,
|
1425 |
+
"loss": 2.8228,
|
1426 |
+
"theoretical_loss": 3.8128981349797098,
|
1427 |
+
"tokens_seen": 642252800
|
1428 |
+
},
|
1429 |
+
{
|
1430 |
+
"epoch": 0.23,
|
1431 |
+
"learning_rate": 0.0003886301305087951,
|
1432 |
+
"loss": 2.8529,
|
1433 |
+
"theoretical_loss": 3.8108989450749293,
|
1434 |
+
"tokens_seen": 645529600
|
1435 |
+
},
|
1436 |
+
{
|
1437 |
+
"epoch": 0.23,
|
1438 |
+
"learning_rate": 0.0003880390580669567,
|
1439 |
+
"loss": 2.8384,
|
1440 |
+
"theoretical_loss": 3.8089127028126764,
|
1441 |
+
"tokens_seen": 648806400
|
1442 |
+
},
|
1443 |
+
{
|
1444 |
+
"epoch": 0.23,
|
1445 |
+
"learning_rate": 0.0003874479856251182,
|
1446 |
+
"loss": 2.8938,
|
1447 |
+
"theoretical_loss": 3.8069392595785083,
|
1448 |
+
"tokens_seen": 652083200
|
1449 |
+
},
|
1450 |
+
{
|
1451 |
+
"debugging/Self-BLEU-5": 0.5265375629586004,
|
1452 |
+
"debugging/distinct-1-grams": 0.7435820408094715,
|
1453 |
+
"debugging/distinct-2-grams": 0.9558103821233092,
|
1454 |
+
"debugging/entropy-1-grams": 5.931434510687563,
|
1455 |
+
"debugging/entropy-2-grams": 6.886416755326388,
|
1456 |
+
"debugging/length": 521.9230769230769,
|
1457 |
+
"debugging/num_segments": 13,
|
1458 |
+
"epoch": 0.23,
|
1459 |
+
"objective/train/docs_used": 379091,
|
1460 |
+
"objective/train/instantaneous_batch_size": 32,
|
1461 |
+
"objective/train/instantaneous_microbatch_size": 32768,
|
1462 |
+
"objective/train/original_loss": 3.0659255981445312,
|
1463 |
+
"objective/train/theoretical_loss": 3.804978469199669,
|
1464 |
+
"objective/train/tokens_used": 675820000,
|
1465 |
+
"theoretical_loss": 3.804978469199669,
|
1466 |
+
"tokens_seen": 655360000
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 0.23,
|
1470 |
+
"learning_rate": 0.00038685691318327974,
|
1471 |
+
"loss": 2.881,
|
1472 |
+
"theoretical_loss": 3.804978469199669,
|
1473 |
+
"tokens_seen": 655360000
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 0.24,
|
1477 |
+
"learning_rate": 0.00038626584074144126,
|
1478 |
+
"loss": 2.888,
|
1479 |
+
"theoretical_loss": 3.803030187893005,
|
1480 |
+
"tokens_seen": 658636800
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 0.24,
|
1484 |
+
"learning_rate": 0.0003856747682996028,
|
1485 |
+
"loss": 2.8765,
|
1486 |
+
"theoretical_loss": 3.8010942742142415,
|
1487 |
+
"tokens_seen": 661913600
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 0.24,
|
1491 |
+
"learning_rate": 0.0003850836958577643,
|
1492 |
+
"loss": 2.9087,
|
1493 |
+
"theoretical_loss": 3.799170589008585,
|
1494 |
+
"tokens_seen": 665190400
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 0.24,
|
1498 |
+
"learning_rate": 0.0003844926234159259,
|
1499 |
+
"loss": 2.8934,
|
1500 |
+
"theoretical_loss": 3.7972589953626006,
|
1501 |
+
"tokens_seen": 668467200
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 0.24,
|
1505 |
+
"learning_rate": 0.0003839015509740874,
|
1506 |
+
"loss": 2.9056,
|
1507 |
+
"theoretical_loss": 3.795359358557337,
|
1508 |
+
"tokens_seen": 671744000
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 0.24,
|
1512 |
+
"learning_rate": 0.0003833104785322489,
|
1513 |
+
"loss": 2.8333,
|
1514 |
+
"theoretical_loss": 3.79347154602265,
|
1515 |
+
"tokens_seen": 675020800
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 0.24,
|
1519 |
+
"learning_rate": 0.00038271940609041044,
|
1520 |
+
"loss": 2.8867,
|
1521 |
+
"theoretical_loss": 3.7915954272926955,
|
1522 |
+
"tokens_seen": 678297600
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 0.24,
|
1526 |
+
"learning_rate": 0.00038212833364857196,
|
1527 |
+
"loss": 2.8426,
|
1528 |
+
"theoretical_loss": 3.789730873962557,
|
1529 |
+
"tokens_seen": 681574400
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 0.24,
|
1533 |
+
"learning_rate": 0.0003815372612067335,
|
1534 |
+
"loss": 2.788,
|
1535 |
+
"theoretical_loss": 3.787877759645963,
|
1536 |
+
"tokens_seen": 684851200
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 0.25,
|
1540 |
+
"learning_rate": 0.000380946188764895,
|
1541 |
+
"loss": 2.8154,
|
1542 |
+
"theoretical_loss": 3.7860359599340776,
|
1543 |
+
"tokens_seen": 688128000
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 0.25,
|
1547 |
+
"learning_rate": 0.0003803551163230566,
|
1548 |
+
"loss": 2.8512,
|
1549 |
+
"theoretical_loss": 3.784205352355321,
|
1550 |
+
"tokens_seen": 691404800
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 0.25,
|
1554 |
+
"learning_rate": 0.0003797640438812181,
|
1555 |
+
"loss": 2.8795,
|
1556 |
+
"theoretical_loss": 3.782385816336189,
|
1557 |
+
"tokens_seen": 694681600
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 0.25,
|
1561 |
+
"learning_rate": 0.0003791729714393796,
|
1562 |
+
"loss": 2.8301,
|
1563 |
+
"theoretical_loss": 3.7805772331630516,
|
1564 |
+
"tokens_seen": 697958400
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 0.25,
|
1568 |
+
"learning_rate": 0.00037858189899754114,
|
1569 |
+
"loss": 2.8087,
|
1570 |
+
"theoretical_loss": 3.7787794859448898,
|
1571 |
+
"tokens_seen": 701235200
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 0.25,
|
1575 |
+
"learning_rate": 0.00037799082655570266,
|
1576 |
+
"loss": 2.7913,
|
1577 |
+
"theoretical_loss": 3.7769924595769546,
|
1578 |
+
"tokens_seen": 704512000
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 0.25,
|
1582 |
+
"learning_rate": 0.0003773997541138642,
|
1583 |
+
"loss": 2.788,
|
1584 |
+
"theoretical_loss": 3.7752160407053115,
|
1585 |
+
"tokens_seen": 707788800
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 0.25,
|
1589 |
+
"learning_rate": 0.00037680868167202575,
|
1590 |
+
"loss": 2.7631,
|
1591 |
+
"theoretical_loss": 3.7734501176922493,
|
1592 |
+
"tokens_seen": 711065600
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 0.26,
|
1596 |
+
"learning_rate": 0.0003762176092301873,
|
1597 |
+
"loss": 2.8143,
|
1598 |
+
"theoretical_loss": 3.7716945805825337,
|
1599 |
+
"tokens_seen": 714342400
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 0.26,
|
1603 |
+
"learning_rate": 0.0003756265367883488,
|
1604 |
+
"loss": 2.8485,
|
1605 |
+
"theoretical_loss": 3.7699493210704667,
|
1606 |
+
"tokens_seen": 717619200
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 0.26,
|
1610 |
+
"learning_rate": 0.0003750354643465103,
|
1611 |
+
"loss": 2.8257,
|
1612 |
+
"theoretical_loss": 3.7682142324677455,
|
1613 |
+
"tokens_seen": 720896000
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 0.26,
|
1617 |
+
"learning_rate": 0.00037444439190467184,
|
1618 |
+
"loss": 2.8877,
|
1619 |
+
"theoretical_loss": 3.7664892096720886,
|
1620 |
+
"tokens_seen": 724172800
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 0.26,
|
1624 |
+
"learning_rate": 0.00037385331946283336,
|
1625 |
+
"loss": 2.8216,
|
1626 |
+
"theoretical_loss": 3.7647741491366067,
|
1627 |
+
"tokens_seen": 727449600
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 0.26,
|
1631 |
+
"learning_rate": 0.00037326224702099493,
|
1632 |
+
"loss": 2.8345,
|
1633 |
+
"theoretical_loss": 3.7630689488399027,
|
1634 |
+
"tokens_seen": 730726400
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 0.26,
|
1638 |
+
"learning_rate": 0.00037267117457915645,
|
1639 |
+
"loss": 2.8008,
|
1640 |
+
"theoretical_loss": 3.7613735082568764,
|
1641 |
+
"tokens_seen": 734003200
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 0.26,
|
1645 |
+
"learning_rate": 0.000372080102137318,
|
1646 |
+
"loss": 2.8302,
|
1647 |
+
"theoretical_loss": 3.759687728330217,
|
1648 |
+
"tokens_seen": 737280000
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 0.26,
|
1652 |
+
"learning_rate": 0.0003714890296954795,
|
1653 |
+
"loss": 2.8284,
|
1654 |
+
"theoretical_loss": 3.75801151144256,
|
1655 |
+
"tokens_seen": 740556800
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 0.27,
|
1659 |
+
"learning_rate": 0.000370897957253641,
|
1660 |
+
"loss": 2.8069,
|
1661 |
+
"theoretical_loss": 3.756344761389295,
|
1662 |
+
"tokens_seen": 743833600
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 0.27,
|
1666 |
+
"learning_rate": 0.00037030688481180254,
|
1667 |
+
"loss": 2.7824,
|
1668 |
+
"theoretical_loss": 3.754687383352003,
|
1669 |
+
"tokens_seen": 747110400
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 0.27,
|
1673 |
+
"learning_rate": 0.00036971581236996406,
|
1674 |
+
"loss": 2.7891,
|
1675 |
+
"theoretical_loss": 3.7530392838725097,
|
1676 |
+
"tokens_seen": 750387200
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 0.27,
|
1680 |
+
"learning_rate": 0.00036912473992812563,
|
1681 |
+
"loss": 2.8107,
|
1682 |
+
"theoretical_loss": 3.751400370827529,
|
1683 |
+
"tokens_seen": 753664000
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 0.27,
|
1687 |
+
"learning_rate": 0.00036853366748628715,
|
1688 |
+
"loss": 2.8383,
|
1689 |
+
"theoretical_loss": 3.749770553403895,
|
1690 |
+
"tokens_seen": 756940800
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 0.27,
|
1694 |
+
"learning_rate": 0.0003679425950444487,
|
1695 |
+
"loss": 2.7993,
|
1696 |
+
"theoretical_loss": 3.748149742074355,
|
1697 |
+
"tokens_seen": 760217600
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 0.27,
|
1701 |
+
"learning_rate": 0.0003673515226026102,
|
1702 |
+
"loss": 2.8346,
|
1703 |
+
"theoretical_loss": 3.746537848573908,
|
1704 |
+
"tokens_seen": 763494400
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 0.27,
|
1708 |
+
"learning_rate": 0.0003667604501607717,
|
1709 |
+
"loss": 2.8492,
|
1710 |
+
"theoretical_loss": 3.744934785876686,
|
1711 |
+
"tokens_seen": 766771200
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 0.28,
|
1715 |
+
"learning_rate": 0.00036616937771893324,
|
1716 |
+
"loss": 2.8489,
|
1717 |
+
"theoretical_loss": 3.7433404681733475,
|
1718 |
+
"tokens_seen": 770048000
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 0.28,
|
1722 |
+
"learning_rate": 0.0003655783052770948,
|
1723 |
+
"loss": 2.8544,
|
1724 |
+
"theoretical_loss": 3.7417548108489846,
|
1725 |
+
"tokens_seen": 773324800
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 0.28,
|
1729 |
+
"learning_rate": 0.00036498723283525633,
|
1730 |
+
"loss": 2.8264,
|
1731 |
+
"theoretical_loss": 3.740177730461517,
|
1732 |
+
"tokens_seen": 776601600
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 0.28,
|
1736 |
+
"learning_rate": 0.00036439616039341785,
|
1737 |
+
"loss": 2.8743,
|
1738 |
+
"theoretical_loss": 3.73860914472057,
|
1739 |
+
"tokens_seen": 779878400
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 0.28,
|
1743 |
+
"learning_rate": 0.0003638050879515794,
|
1744 |
+
"loss": 2.8407,
|
1745 |
+
"theoretical_loss": 3.7370489724668197,
|
1746 |
+
"tokens_seen": 783155200
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 0.28,
|
1750 |
+
"learning_rate": 0.0003632140155097409,
|
1751 |
+
"loss": 2.8535,
|
1752 |
+
"theoretical_loss": 3.735497133651788,
|
1753 |
+
"tokens_seen": 786432000
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 0.28,
|
1757 |
+
"learning_rate": 0.0003626229430679024,
|
1758 |
+
"loss": 2.836,
|
1759 |
+
"theoretical_loss": 3.733953549318091,
|
1760 |
+
"tokens_seen": 789708800
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 0.28,
|
1764 |
+
"learning_rate": 0.000362031870626064,
|
1765 |
+
"loss": 2.8547,
|
1766 |
+
"theoretical_loss": 3.7324181415801094,
|
1767 |
+
"tokens_seen": 792985600
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 0.28,
|
1771 |
+
"learning_rate": 0.00036144079818422546,
|
1772 |
+
"loss": 2.8486,
|
1773 |
+
"theoretical_loss": 3.7308908336050814,
|
1774 |
+
"tokens_seen": 796262400
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 0.29,
|
1778 |
+
"learning_rate": 0.000360849725742387,
|
1779 |
+
"loss": 2.8293,
|
1780 |
+
"theoretical_loss": 3.729371549594614,
|
1781 |
+
"tokens_seen": 799539200
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 0.29,
|
1785 |
+
"learning_rate": 0.0003602586533005485,
|
1786 |
+
"loss": 2.8396,
|
1787 |
+
"theoretical_loss": 3.7278602147665776,
|
1788 |
+
"tokens_seen": 802816000
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 0.29,
|
1792 |
+
"learning_rate": 0.00035966758085871,
|
1793 |
+
"loss": 2.8218,
|
1794 |
+
"theoretical_loss": 3.726356755337407,
|
1795 |
+
"tokens_seen": 806092800
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 0.29,
|
1799 |
+
"learning_rate": 0.00035907650841687154,
|
1800 |
+
"loss": 2.8109,
|
1801 |
+
"theoretical_loss": 3.724861098504767,
|
1802 |
+
"tokens_seen": 809369600
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 0.29,
|
1806 |
+
"learning_rate": 0.0003584972574238699,
|
1807 |
+
"loss": 2.7916,
|
1808 |
+
"theoretical_loss": 3.7233731724305974,
|
1809 |
+
"tokens_seen": 812646400
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 0.29,
|
1813 |
+
"learning_rate": 0.00035790618498203144,
|
1814 |
+
"loss": 2.8143,
|
1815 |
+
"theoretical_loss": 3.7218929062245105,
|
1816 |
+
"tokens_seen": 815923200
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 0.29,
|
1820 |
+
"objective/train/docs_used": 471128,
|
1821 |
+
"objective/train/instantaneous_batch_size": 32,
|
1822 |
+
"objective/train/instantaneous_microbatch_size": 32768,
|
1823 |
+
"objective/train/original_loss": 2.6718621253967285,
|
1824 |
+
"objective/train/theoretical_loss": 3.7204202299275475,
|
1825 |
+
"objective/train/tokens_used": 839660000,
|
1826 |
+
"theoretical_loss": 3.7204202299275475,
|
1827 |
+
"tokens_seen": 819200000
|
1828 |
+
},
|
1829 |
+
{
|
1830 |
+
"epoch": 0.29,
|
1831 |
+
"learning_rate": 0.00035731511254019296,
|
1832 |
+
"loss": 2.7586,
|
1833 |
+
"theoretical_loss": 3.7204202299275475,
|
1834 |
+
"tokens_seen": 819200000
|
1835 |
+
},
|
1836 |
+
{
|
1837 |
+
"epoch": 0.29,
|
1838 |
+
"learning_rate": 0.0003567240400983545,
|
1839 |
+
"loss": 2.78,
|
1840 |
+
"theoretical_loss": 3.7189550744962707,
|
1841 |
+
"tokens_seen": 822476800
|
1842 |
+
},
|
1843 |
+
{
|
1844 |
+
"epoch": 0.29,
|
1845 |
+
"learning_rate": 0.000356132967656516,
|
1846 |
+
"loss": 2.762,
|
1847 |
+
"theoretical_loss": 3.717497371787192,
|
1848 |
+
"tokens_seen": 825753600
|
1849 |
+
},
|
1850 |
+
{
|
1851 |
+
"epoch": 0.3,
|
1852 |
+
"learning_rate": 0.0003555418952146775,
|
1853 |
+
"loss": 2.7573,
|
1854 |
+
"theoretical_loss": 3.7160470545415274,
|
1855 |
+
"tokens_seen": 829030400
|
1856 |
+
},
|
1857 |
+
{
|
1858 |
+
"epoch": 0.3,
|
1859 |
+
"learning_rate": 0.00035495082277283904,
|
1860 |
+
"loss": 2.7581,
|
1861 |
+
"theoretical_loss": 3.714604056370267,
|
1862 |
+
"tokens_seen": 832307200
|
1863 |
+
},
|
1864 |
+
{
|
1865 |
+
"epoch": 0.3,
|
1866 |
+
"learning_rate": 0.0003543597503310006,
|
1867 |
+
"loss": 2.7728,
|
1868 |
+
"theoretical_loss": 3.713168311739558,
|
1869 |
+
"tokens_seen": 835584000
|
1870 |
+
},
|
1871 |
+
{
|
1872 |
+
"epoch": 0.3,
|
1873 |
+
"learning_rate": 0.00035376867788916214,
|
1874 |
+
"loss": 2.7538,
|
1875 |
+
"theoretical_loss": 3.7117397559563843,
|
1876 |
+
"tokens_seen": 838860800
|
1877 |
+
}
|
1878 |
+
],
|
1879 |
+
"max_steps": 42724,
|
1880 |
+
"num_train_epochs": 9223372036854775807,
|
1881 |
+
"total_flos": 4.28635991310336e+17,
|
1882 |
+
"trial_name": null,
|
1883 |
+
"trial_params": null
|
1884 |
+
}
|
checkpoint-12816/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b99f0235a96529e0f83ff582cf9a79155af5321a32b48f56f5aaa8415957f796
|
3 |
+
size 3451
|
checkpoint-12816/vocab.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 510398013
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3bc4cf094124a3fd9b78471cdeec9c812040347ceebfa4c5c137d8a18ca6c37d
|
3 |
size 510398013
|