FastPlus125m Geodesic Backup checkpoint-2000 at global step 2000
Browse files- checkpoint-2000/config.json +19 -0
- checkpoint-2000/generation_config.json +6 -0
- checkpoint-2000/model.safetensors +3 -0
- checkpoint-2000/optimizer.pt +3 -0
- checkpoint-2000/rng_state.pth +3 -0
- checkpoint-2000/scaler.pt +3 -0
- checkpoint-2000/scheduler.pt +3 -0
- checkpoint-2000/tokenizer.json +0 -0
- checkpoint-2000/tokenizer_config.json +15 -0
- checkpoint-2000/trainer_state.json +1634 -0
- checkpoint-2000/training_args.bin +3 -0
checkpoint-2000/config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"FastPlus125mForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"dtype": "float32",
|
| 6 |
+
"hidden_size": 768,
|
| 7 |
+
"initializer_range": 0.02,
|
| 8 |
+
"intermediate_size": 2048,
|
| 9 |
+
"kd_alpha": 0.4,
|
| 10 |
+
"kd_temperature": 2.5,
|
| 11 |
+
"max_position_embeddings": 512,
|
| 12 |
+
"model_type": "fastplus_125m",
|
| 13 |
+
"num_attention_heads": 12,
|
| 14 |
+
"num_hidden_layers": 16,
|
| 15 |
+
"tie_word_embeddings": false,
|
| 16 |
+
"transformers_version": "5.12.0",
|
| 17 |
+
"use_cache": false,
|
| 18 |
+
"vocab_size": 32704
|
| 19 |
+
}
|
checkpoint-2000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"output_attentions": false,
|
| 4 |
+
"output_hidden_states": false,
|
| 5 |
+
"transformers_version": "5.12.0"
|
| 6 |
+
}
|
checkpoint-2000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:093288f944c52f67c691db97f4d06795eae4f0cdec00b49b6b583ced4f83da35
|
| 3 |
+
size 573839072
|
checkpoint-2000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4fcf6b42681737a251f21457ee8ec282dc48c6017b1954f6981a935b243ec942
|
| 3 |
+
size 1140607051
|
checkpoint-2000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61c19bab1174704a4a4441475683bf1270277af15d2e2c95e964789128e482c4
|
| 3 |
+
size 14645
|
checkpoint-2000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df8cef26aac1f8b2fdbc493100ca72702ece53b58e7a93ad3420fc3a5dac68bb
|
| 3 |
+
size 1383
|
checkpoint-2000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a89fdf2912ff6891fbf49a2bbac6d18730e0e4b2e34d5ceddd76c51b38baf55d
|
| 3 |
+
size 7737
|
checkpoint-2000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"extra_special_tokens": [
|
| 6 |
+
"<|user|>",
|
| 7 |
+
"<|assistant|>"
|
| 8 |
+
],
|
| 9 |
+
"is_local": false,
|
| 10 |
+
"local_files_only": false,
|
| 11 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"tokenizer_class": "TokenizersBackend",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
checkpoint-2000/trainer_state.json
ADDED
|
@@ -0,0 +1,1634 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.032237915818742316,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0001611895790937116,
|
| 14 |
+
"grad_norm": Infinity,
|
| 15 |
+
"learning_rate": 1.7000000000000001e-07,
|
| 16 |
+
"lorentz_dist_loss": 1.1774,
|
| 17 |
+
"loss": 14.170442199707031,
|
| 18 |
+
"step": 10
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.0003223791581874232,
|
| 22 |
+
"grad_norm": 11644.1904296875,
|
| 23 |
+
"learning_rate": 6.800000000000001e-07,
|
| 24 |
+
"lorentz_dist_loss": 1.1762,
|
| 25 |
+
"loss": 14.155226135253907,
|
| 26 |
+
"step": 20
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.0004835687372811348,
|
| 30 |
+
"grad_norm": 11821.7451171875,
|
| 31 |
+
"learning_rate": 1.19e-06,
|
| 32 |
+
"lorentz_dist_loss": 1.166,
|
| 33 |
+
"loss": 14.033517456054687,
|
| 34 |
+
"step": 30
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.0006447583163748464,
|
| 38 |
+
"grad_norm": 10725.34375,
|
| 39 |
+
"learning_rate": 1.7566666666666669e-06,
|
| 40 |
+
"lorentz_dist_loss": 1.1524,
|
| 41 |
+
"loss": 13.870587158203126,
|
| 42 |
+
"step": 40
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.0008059478954685579,
|
| 46 |
+
"grad_norm": 10575.232421875,
|
| 47 |
+
"learning_rate": 2.3233333333333337e-06,
|
| 48 |
+
"lorentz_dist_loss": 1.1312,
|
| 49 |
+
"loss": 13.616365051269531,
|
| 50 |
+
"step": 50
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.0009671374745622695,
|
| 54 |
+
"grad_norm": 10407.58984375,
|
| 55 |
+
"learning_rate": 2.8900000000000003e-06,
|
| 56 |
+
"lorentz_dist_loss": 1.1051,
|
| 57 |
+
"loss": 13.303155517578125,
|
| 58 |
+
"step": 60
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.001128327053655981,
|
| 62 |
+
"grad_norm": 9515.619140625,
|
| 63 |
+
"learning_rate": 3.4566666666666665e-06,
|
| 64 |
+
"lorentz_dist_loss": 1.0727,
|
| 65 |
+
"loss": 12.913534545898438,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.0012895166327496928,
|
| 70 |
+
"grad_norm": 8322.388671875,
|
| 71 |
+
"learning_rate": 4.0233333333333335e-06,
|
| 72 |
+
"lorentz_dist_loss": 1.0396,
|
| 73 |
+
"loss": 12.51680908203125,
|
| 74 |
+
"step": 80
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.0014507062118434043,
|
| 78 |
+
"grad_norm": 7218.798828125,
|
| 79 |
+
"learning_rate": 4.59e-06,
|
| 80 |
+
"lorentz_dist_loss": 1.0072,
|
| 81 |
+
"loss": 12.128275299072266,
|
| 82 |
+
"step": 90
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.0016118957909371158,
|
| 86 |
+
"grad_norm": 5936.27294921875,
|
| 87 |
+
"learning_rate": 5.156666666666667e-06,
|
| 88 |
+
"lorentz_dist_loss": 0.9771,
|
| 89 |
+
"loss": 11.766087341308594,
|
| 90 |
+
"step": 100
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.0017730853700308276,
|
| 94 |
+
"grad_norm": 4860.89990234375,
|
| 95 |
+
"learning_rate": 5.723333333333333e-06,
|
| 96 |
+
"lorentz_dist_loss": 0.9544,
|
| 97 |
+
"loss": 11.49432144165039,
|
| 98 |
+
"step": 110
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.001934274949124539,
|
| 102 |
+
"grad_norm": 3848.34814453125,
|
| 103 |
+
"learning_rate": 6.29e-06,
|
| 104 |
+
"lorentz_dist_loss": 0.9391,
|
| 105 |
+
"loss": 11.310404205322266,
|
| 106 |
+
"step": 120
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.002095464528218251,
|
| 110 |
+
"grad_norm": 3314.070068359375,
|
| 111 |
+
"learning_rate": 6.856666666666667e-06,
|
| 112 |
+
"lorentz_dist_loss": 0.9293,
|
| 113 |
+
"loss": 11.192581939697266,
|
| 114 |
+
"step": 130
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.002256654107311962,
|
| 118 |
+
"grad_norm": 2885.541015625,
|
| 119 |
+
"learning_rate": 7.423333333333334e-06,
|
| 120 |
+
"lorentz_dist_loss": 0.9242,
|
| 121 |
+
"loss": 11.131652069091796,
|
| 122 |
+
"step": 140
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.002417843686405674,
|
| 126 |
+
"grad_norm": 2752.97021484375,
|
| 127 |
+
"learning_rate": 7.990000000000001e-06,
|
| 128 |
+
"lorentz_dist_loss": 0.9203,
|
| 129 |
+
"loss": 11.085308837890626,
|
| 130 |
+
"step": 150
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.0025790332654993856,
|
| 134 |
+
"grad_norm": 2966.74853515625,
|
| 135 |
+
"learning_rate": 8.556666666666667e-06,
|
| 136 |
+
"lorentz_dist_loss": 0.9166,
|
| 137 |
+
"loss": 11.040792083740234,
|
| 138 |
+
"step": 160
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.002740222844593097,
|
| 142 |
+
"grad_norm": 2726.117919921875,
|
| 143 |
+
"learning_rate": 9.123333333333335e-06,
|
| 144 |
+
"lorentz_dist_loss": 0.9137,
|
| 145 |
+
"loss": 11.00632095336914,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.0029014124236868086,
|
| 150 |
+
"grad_norm": 2733.96728515625,
|
| 151 |
+
"learning_rate": 9.69e-06,
|
| 152 |
+
"lorentz_dist_loss": 0.9113,
|
| 153 |
+
"loss": 10.977420043945312,
|
| 154 |
+
"step": 180
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.0030626020027805204,
|
| 158 |
+
"grad_norm": 2651.473876953125,
|
| 159 |
+
"learning_rate": 1.0256666666666668e-05,
|
| 160 |
+
"lorentz_dist_loss": 0.9076,
|
| 161 |
+
"loss": 10.932640838623048,
|
| 162 |
+
"step": 190
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.0032237915818742317,
|
| 166 |
+
"grad_norm": 2543.915771484375,
|
| 167 |
+
"learning_rate": 1.0823333333333334e-05,
|
| 168 |
+
"lorentz_dist_loss": 0.9051,
|
| 169 |
+
"loss": 10.90311279296875,
|
| 170 |
+
"step": 200
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.0033849811609679434,
|
| 174 |
+
"grad_norm": 2337.19921875,
|
| 175 |
+
"learning_rate": 1.1390000000000001e-05,
|
| 176 |
+
"lorentz_dist_loss": 0.9029,
|
| 177 |
+
"loss": 10.87609634399414,
|
| 178 |
+
"step": 210
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.003546170740061655,
|
| 182 |
+
"grad_norm": 2522.357666015625,
|
| 183 |
+
"learning_rate": 1.1956666666666667e-05,
|
| 184 |
+
"lorentz_dist_loss": 0.9012,
|
| 185 |
+
"loss": 10.855293273925781,
|
| 186 |
+
"step": 220
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.0037073603191553665,
|
| 190 |
+
"grad_norm": 2252.1298828125,
|
| 191 |
+
"learning_rate": 1.2523333333333336e-05,
|
| 192 |
+
"lorentz_dist_loss": 0.8996,
|
| 193 |
+
"loss": 10.836546325683594,
|
| 194 |
+
"step": 230
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.003868549898249078,
|
| 198 |
+
"grad_norm": 1921.8360595703125,
|
| 199 |
+
"learning_rate": 1.309e-05,
|
| 200 |
+
"lorentz_dist_loss": 0.8972,
|
| 201 |
+
"loss": 10.808248901367188,
|
| 202 |
+
"step": 240
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.0040297394773427895,
|
| 206 |
+
"grad_norm": 2205.89794921875,
|
| 207 |
+
"learning_rate": 1.365666666666667e-05,
|
| 208 |
+
"lorentz_dist_loss": 0.8945,
|
| 209 |
+
"loss": 10.774971771240235,
|
| 210 |
+
"step": 250
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.004190929056436502,
|
| 214 |
+
"grad_norm": 1951.3402099609375,
|
| 215 |
+
"learning_rate": 1.4223333333333335e-05,
|
| 216 |
+
"lorentz_dist_loss": 0.8923,
|
| 217 |
+
"loss": 10.748961639404296,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.004352118635530213,
|
| 222 |
+
"grad_norm": 1976.0369873046875,
|
| 223 |
+
"learning_rate": 1.479e-05,
|
| 224 |
+
"lorentz_dist_loss": 0.891,
|
| 225 |
+
"loss": 10.73399658203125,
|
| 226 |
+
"step": 270
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.004513308214623924,
|
| 230 |
+
"grad_norm": 1986.1494140625,
|
| 231 |
+
"learning_rate": 1.5356666666666668e-05,
|
| 232 |
+
"lorentz_dist_loss": 0.8896,
|
| 233 |
+
"loss": 10.7161376953125,
|
| 234 |
+
"step": 280
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.0046744977937176364,
|
| 238 |
+
"grad_norm": 2266.1396484375,
|
| 239 |
+
"learning_rate": 1.5923333333333334e-05,
|
| 240 |
+
"lorentz_dist_loss": 0.8868,
|
| 241 |
+
"loss": 10.683519744873047,
|
| 242 |
+
"step": 290
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.004835687372811348,
|
| 246 |
+
"grad_norm": 1954.756103515625,
|
| 247 |
+
"learning_rate": 1.6490000000000003e-05,
|
| 248 |
+
"lorentz_dist_loss": 0.8857,
|
| 249 |
+
"loss": 10.669470977783202,
|
| 250 |
+
"step": 300
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.004996876951905059,
|
| 254 |
+
"grad_norm": 1940.30517578125,
|
| 255 |
+
"learning_rate": 1.705666666666667e-05,
|
| 256 |
+
"lorentz_dist_loss": 0.8848,
|
| 257 |
+
"loss": 10.659544372558594,
|
| 258 |
+
"step": 310
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.005158066530998771,
|
| 262 |
+
"grad_norm": 1976.4127197265625,
|
| 263 |
+
"learning_rate": 1.7623333333333335e-05,
|
| 264 |
+
"lorentz_dist_loss": 0.8826,
|
| 265 |
+
"loss": 10.632625579833984,
|
| 266 |
+
"step": 320
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.0053192561100924825,
|
| 270 |
+
"grad_norm": 1840.610107421875,
|
| 271 |
+
"learning_rate": 1.819e-05,
|
| 272 |
+
"lorentz_dist_loss": 0.8801,
|
| 273 |
+
"loss": 10.602033233642578,
|
| 274 |
+
"step": 330
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.005480445689186194,
|
| 278 |
+
"grad_norm": 1728.57080078125,
|
| 279 |
+
"learning_rate": 1.875666666666667e-05,
|
| 280 |
+
"lorentz_dist_loss": 0.8781,
|
| 281 |
+
"loss": 10.578665924072265,
|
| 282 |
+
"step": 340
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.005641635268279906,
|
| 286 |
+
"grad_norm": 1930.06396484375,
|
| 287 |
+
"learning_rate": 1.9323333333333335e-05,
|
| 288 |
+
"lorentz_dist_loss": 0.8779,
|
| 289 |
+
"loss": 10.576192474365234,
|
| 290 |
+
"step": 350
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.005802824847373617,
|
| 294 |
+
"grad_norm": 1823.3944091796875,
|
| 295 |
+
"learning_rate": 1.989e-05,
|
| 296 |
+
"lorentz_dist_loss": 0.8759,
|
| 297 |
+
"loss": 10.552652740478516,
|
| 298 |
+
"step": 360
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.005964014426467329,
|
| 302 |
+
"grad_norm": 1744.7955322265625,
|
| 303 |
+
"learning_rate": 2.0456666666666667e-05,
|
| 304 |
+
"lorentz_dist_loss": 0.8744,
|
| 305 |
+
"loss": 10.534652709960938,
|
| 306 |
+
"step": 370
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.006125204005561041,
|
| 310 |
+
"grad_norm": 1688.642822265625,
|
| 311 |
+
"learning_rate": 2.1023333333333333e-05,
|
| 312 |
+
"lorentz_dist_loss": 0.8718,
|
| 313 |
+
"loss": 10.50313720703125,
|
| 314 |
+
"step": 380
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.006286393584654752,
|
| 318 |
+
"grad_norm": 1514.7152099609375,
|
| 319 |
+
"learning_rate": 2.159e-05,
|
| 320 |
+
"lorentz_dist_loss": 0.8718,
|
| 321 |
+
"loss": 10.503563690185548,
|
| 322 |
+
"step": 390
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.006447583163748463,
|
| 326 |
+
"grad_norm": 1295.44287109375,
|
| 327 |
+
"learning_rate": 2.2156666666666667e-05,
|
| 328 |
+
"lorentz_dist_loss": 0.8697,
|
| 329 |
+
"loss": 10.478243255615235,
|
| 330 |
+
"step": 400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.0066087727428421755,
|
| 334 |
+
"grad_norm": 1421.1474609375,
|
| 335 |
+
"learning_rate": 2.2723333333333333e-05,
|
| 336 |
+
"lorentz_dist_loss": 0.869,
|
| 337 |
+
"loss": 10.469520568847656,
|
| 338 |
+
"step": 410
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.006769962321935887,
|
| 342 |
+
"grad_norm": 1567.665283203125,
|
| 343 |
+
"learning_rate": 2.3290000000000002e-05,
|
| 344 |
+
"lorentz_dist_loss": 0.869,
|
| 345 |
+
"loss": 10.469043731689453,
|
| 346 |
+
"step": 420
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.006931151901029598,
|
| 350 |
+
"grad_norm": 1557.412109375,
|
| 351 |
+
"learning_rate": 2.3856666666666668e-05,
|
| 352 |
+
"lorentz_dist_loss": 0.8672,
|
| 353 |
+
"loss": 10.448148345947265,
|
| 354 |
+
"step": 430
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.00709234148012331,
|
| 358 |
+
"grad_norm": 1410.789794921875,
|
| 359 |
+
"learning_rate": 2.4423333333333334e-05,
|
| 360 |
+
"lorentz_dist_loss": 0.8657,
|
| 361 |
+
"loss": 10.429232788085937,
|
| 362 |
+
"step": 440
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.007253531059217022,
|
| 366 |
+
"grad_norm": 1325.1748046875,
|
| 367 |
+
"learning_rate": 2.499e-05,
|
| 368 |
+
"lorentz_dist_loss": 0.8648,
|
| 369 |
+
"loss": 10.418812561035157,
|
| 370 |
+
"step": 450
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.007414720638310733,
|
| 374 |
+
"grad_norm": 1519.82421875,
|
| 375 |
+
"learning_rate": 2.5556666666666672e-05,
|
| 376 |
+
"lorentz_dist_loss": 0.8645,
|
| 377 |
+
"loss": 10.415811920166016,
|
| 378 |
+
"step": 460
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.007575910217404445,
|
| 382 |
+
"grad_norm": 1237.5809326171875,
|
| 383 |
+
"learning_rate": 2.6123333333333338e-05,
|
| 384 |
+
"lorentz_dist_loss": 0.8615,
|
| 385 |
+
"loss": 10.378885650634766,
|
| 386 |
+
"step": 470
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.007737099796498156,
|
| 390 |
+
"grad_norm": 1281.034423828125,
|
| 391 |
+
"learning_rate": 2.6690000000000004e-05,
|
| 392 |
+
"lorentz_dist_loss": 0.8607,
|
| 393 |
+
"loss": 10.370066833496093,
|
| 394 |
+
"step": 480
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.007898289375591868,
|
| 398 |
+
"grad_norm": 1342.318603515625,
|
| 399 |
+
"learning_rate": 2.7256666666666666e-05,
|
| 400 |
+
"lorentz_dist_loss": 0.8604,
|
| 401 |
+
"loss": 10.365609741210937,
|
| 402 |
+
"step": 490
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.008059478954685579,
|
| 406 |
+
"grad_norm": 1032.6192626953125,
|
| 407 |
+
"learning_rate": 2.7823333333333332e-05,
|
| 408 |
+
"lorentz_dist_loss": 0.8587,
|
| 409 |
+
"loss": 10.346303558349609,
|
| 410 |
+
"step": 500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.008220668533779292,
|
| 414 |
+
"grad_norm": 1070.57421875,
|
| 415 |
+
"learning_rate": 2.8390000000000004e-05,
|
| 416 |
+
"lorentz_dist_loss": 0.8581,
|
| 417 |
+
"loss": 10.338904571533202,
|
| 418 |
+
"step": 510
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.008381858112873003,
|
| 422 |
+
"grad_norm": 1049.3587646484375,
|
| 423 |
+
"learning_rate": 2.895666666666667e-05,
|
| 424 |
+
"lorentz_dist_loss": 0.8561,
|
| 425 |
+
"loss": 10.314006805419922,
|
| 426 |
+
"step": 520
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.008543047691966715,
|
| 430 |
+
"grad_norm": 899.3391723632812,
|
| 431 |
+
"learning_rate": 2.9523333333333336e-05,
|
| 432 |
+
"lorentz_dist_loss": 0.8544,
|
| 433 |
+
"loss": 10.294047546386718,
|
| 434 |
+
"step": 530
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.008704237271060426,
|
| 438 |
+
"grad_norm": 1103.1273193359375,
|
| 439 |
+
"learning_rate": 3.009e-05,
|
| 440 |
+
"lorentz_dist_loss": 0.8545,
|
| 441 |
+
"loss": 10.294941711425782,
|
| 442 |
+
"step": 540
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.008865426850154137,
|
| 446 |
+
"grad_norm": 816.0887451171875,
|
| 447 |
+
"learning_rate": 3.0656666666666674e-05,
|
| 448 |
+
"lorentz_dist_loss": 0.8527,
|
| 449 |
+
"loss": 10.273906707763672,
|
| 450 |
+
"step": 550
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.009026616429247849,
|
| 454 |
+
"grad_norm": 789.2012939453125,
|
| 455 |
+
"learning_rate": 3.1223333333333336e-05,
|
| 456 |
+
"lorentz_dist_loss": 0.8517,
|
| 457 |
+
"loss": 10.262254333496093,
|
| 458 |
+
"step": 560
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.009187806008341562,
|
| 462 |
+
"grad_norm": 653.559326171875,
|
| 463 |
+
"learning_rate": 3.179e-05,
|
| 464 |
+
"lorentz_dist_loss": 0.8522,
|
| 465 |
+
"loss": 10.26751251220703,
|
| 466 |
+
"step": 570
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.009348995587435273,
|
| 470 |
+
"grad_norm": 691.96337890625,
|
| 471 |
+
"learning_rate": 3.235666666666667e-05,
|
| 472 |
+
"lorentz_dist_loss": 0.8496,
|
| 473 |
+
"loss": 10.236648559570312,
|
| 474 |
+
"step": 580
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.009510185166528984,
|
| 478 |
+
"grad_norm": 613.22802734375,
|
| 479 |
+
"learning_rate": 3.292333333333333e-05,
|
| 480 |
+
"lorentz_dist_loss": 0.8491,
|
| 481 |
+
"loss": 10.229987335205077,
|
| 482 |
+
"step": 590
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.009671374745622695,
|
| 486 |
+
"grad_norm": 667.1336669921875,
|
| 487 |
+
"learning_rate": 3.3490000000000006e-05,
|
| 488 |
+
"lorentz_dist_loss": 0.8489,
|
| 489 |
+
"loss": 10.228546142578125,
|
| 490 |
+
"step": 600
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.009832564324716407,
|
| 494 |
+
"grad_norm": 575.3295288085938,
|
| 495 |
+
"learning_rate": 3.405666666666667e-05,
|
| 496 |
+
"lorentz_dist_loss": 0.8474,
|
| 497 |
+
"loss": 10.209774780273438,
|
| 498 |
+
"step": 610
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.009993753903810118,
|
| 502 |
+
"grad_norm": 566.0633544921875,
|
| 503 |
+
"learning_rate": 3.462333333333334e-05,
|
| 504 |
+
"lorentz_dist_loss": 0.8457,
|
| 505 |
+
"loss": 10.190084075927734,
|
| 506 |
+
"step": 620
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.010154943482903831,
|
| 510 |
+
"grad_norm": 518.58740234375,
|
| 511 |
+
"learning_rate": 3.519e-05,
|
| 512 |
+
"lorentz_dist_loss": 0.8468,
|
| 513 |
+
"loss": 10.202525329589843,
|
| 514 |
+
"step": 630
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.010316133061997542,
|
| 518 |
+
"grad_norm": 533.7039184570312,
|
| 519 |
+
"learning_rate": 3.575666666666667e-05,
|
| 520 |
+
"lorentz_dist_loss": 0.8466,
|
| 521 |
+
"loss": 10.199913024902344,
|
| 522 |
+
"step": 640
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.010477322641091254,
|
| 526 |
+
"grad_norm": 479.345703125,
|
| 527 |
+
"learning_rate": 3.632333333333334e-05,
|
| 528 |
+
"lorentz_dist_loss": 0.8436,
|
| 529 |
+
"loss": 10.164936065673828,
|
| 530 |
+
"step": 650
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.010638512220184965,
|
| 534 |
+
"grad_norm": 458.440673828125,
|
| 535 |
+
"learning_rate": 3.689e-05,
|
| 536 |
+
"lorentz_dist_loss": 0.8442,
|
| 537 |
+
"loss": 10.171356201171875,
|
| 538 |
+
"step": 660
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 0.010799701799278676,
|
| 542 |
+
"grad_norm": 453.0855712890625,
|
| 543 |
+
"learning_rate": 3.745666666666667e-05,
|
| 544 |
+
"lorentz_dist_loss": 0.8435,
|
| 545 |
+
"loss": 10.163714599609374,
|
| 546 |
+
"step": 670
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 0.010960891378372388,
|
| 550 |
+
"grad_norm": 463.44342041015625,
|
| 551 |
+
"learning_rate": 3.802333333333334e-05,
|
| 552 |
+
"lorentz_dist_loss": 0.8436,
|
| 553 |
+
"loss": 10.164224243164062,
|
| 554 |
+
"step": 680
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.0111220809574661,
|
| 558 |
+
"grad_norm": 511.4906921386719,
|
| 559 |
+
"learning_rate": 3.859e-05,
|
| 560 |
+
"lorentz_dist_loss": 0.8432,
|
| 561 |
+
"loss": 10.159163665771484,
|
| 562 |
+
"step": 690
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.011283270536559812,
|
| 566 |
+
"grad_norm": 538.5718994140625,
|
| 567 |
+
"learning_rate": 3.915666666666667e-05,
|
| 568 |
+
"lorentz_dist_loss": 0.8419,
|
| 569 |
+
"loss": 10.144430541992188,
|
| 570 |
+
"step": 700
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.011444460115653523,
|
| 574 |
+
"grad_norm": 465.85638427734375,
|
| 575 |
+
"learning_rate": 3.972333333333333e-05,
|
| 576 |
+
"lorentz_dist_loss": 0.8418,
|
| 577 |
+
"loss": 10.143357086181641,
|
| 578 |
+
"step": 710
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.011605649694747235,
|
| 582 |
+
"grad_norm": 502.9533996582031,
|
| 583 |
+
"learning_rate": 4.029e-05,
|
| 584 |
+
"lorentz_dist_loss": 0.8396,
|
| 585 |
+
"loss": 10.115806579589844,
|
| 586 |
+
"step": 720
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.011766839273840946,
|
| 590 |
+
"grad_norm": 527.615478515625,
|
| 591 |
+
"learning_rate": 4.085666666666667e-05,
|
| 592 |
+
"lorentz_dist_loss": 0.8411,
|
| 593 |
+
"loss": 10.134616088867187,
|
| 594 |
+
"step": 730
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.011928028852934657,
|
| 598 |
+
"grad_norm": 510.4620361328125,
|
| 599 |
+
"learning_rate": 4.1423333333333334e-05,
|
| 600 |
+
"lorentz_dist_loss": 0.8405,
|
| 601 |
+
"loss": 10.126832580566406,
|
| 602 |
+
"step": 740
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.01208921843202837,
|
| 606 |
+
"grad_norm": 482.1481628417969,
|
| 607 |
+
"learning_rate": 4.199e-05,
|
| 608 |
+
"lorentz_dist_loss": 0.8398,
|
| 609 |
+
"loss": 10.118888092041015,
|
| 610 |
+
"step": 750
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.012250408011122082,
|
| 614 |
+
"grad_norm": 483.4075927734375,
|
| 615 |
+
"learning_rate": 4.255666666666667e-05,
|
| 616 |
+
"lorentz_dist_loss": 0.8384,
|
| 617 |
+
"loss": 10.102037811279297,
|
| 618 |
+
"step": 760
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.012411597590215793,
|
| 622 |
+
"grad_norm": 524.883056640625,
|
| 623 |
+
"learning_rate": 4.3123333333333334e-05,
|
| 624 |
+
"lorentz_dist_loss": 0.8388,
|
| 625 |
+
"loss": 10.107166290283203,
|
| 626 |
+
"step": 770
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.012572787169309504,
|
| 630 |
+
"grad_norm": 475.61236572265625,
|
| 631 |
+
"learning_rate": 4.3690000000000004e-05,
|
| 632 |
+
"lorentz_dist_loss": 0.8374,
|
| 633 |
+
"loss": 10.089569854736329,
|
| 634 |
+
"step": 780
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.012733976748403215,
|
| 638 |
+
"grad_norm": 520.844970703125,
|
| 639 |
+
"learning_rate": 4.425666666666667e-05,
|
| 640 |
+
"lorentz_dist_loss": 0.8382,
|
| 641 |
+
"loss": 10.099609375,
|
| 642 |
+
"step": 790
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.012895166327496927,
|
| 646 |
+
"grad_norm": 602.2792358398438,
|
| 647 |
+
"learning_rate": 4.4823333333333335e-05,
|
| 648 |
+
"lorentz_dist_loss": 0.838,
|
| 649 |
+
"loss": 10.097400665283203,
|
| 650 |
+
"step": 800
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.01305635590659064,
|
| 654 |
+
"grad_norm": 441.967529296875,
|
| 655 |
+
"learning_rate": 4.5390000000000004e-05,
|
| 656 |
+
"lorentz_dist_loss": 0.8367,
|
| 657 |
+
"loss": 10.081182098388672,
|
| 658 |
+
"step": 810
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 0.013217545485684351,
|
| 662 |
+
"grad_norm": 448.0221252441406,
|
| 663 |
+
"learning_rate": 4.5956666666666667e-05,
|
| 664 |
+
"lorentz_dist_loss": 0.8383,
|
| 665 |
+
"loss": 10.100080871582032,
|
| 666 |
+
"step": 820
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.013378735064778062,
|
| 670 |
+
"grad_norm": 543.4454345703125,
|
| 671 |
+
"learning_rate": 4.6523333333333336e-05,
|
| 672 |
+
"lorentz_dist_loss": 0.8367,
|
| 673 |
+
"loss": 10.081255340576172,
|
| 674 |
+
"step": 830
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.013539924643871774,
|
| 678 |
+
"grad_norm": 482.30706787109375,
|
| 679 |
+
"learning_rate": 4.7090000000000005e-05,
|
| 680 |
+
"lorentz_dist_loss": 0.8359,
|
| 681 |
+
"loss": 10.071501922607421,
|
| 682 |
+
"step": 840
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.013701114222965485,
|
| 686 |
+
"grad_norm": 505.6029052734375,
|
| 687 |
+
"learning_rate": 4.765666666666667e-05,
|
| 688 |
+
"lorentz_dist_loss": 0.8347,
|
| 689 |
+
"loss": 10.057789611816407,
|
| 690 |
+
"step": 850
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.013862303802059196,
|
| 694 |
+
"grad_norm": 483.9949951171875,
|
| 695 |
+
"learning_rate": 4.8223333333333336e-05,
|
| 696 |
+
"lorentz_dist_loss": 0.8344,
|
| 697 |
+
"loss": 10.05423812866211,
|
| 698 |
+
"step": 860
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.01402349338115291,
|
| 702 |
+
"grad_norm": 513.040771484375,
|
| 703 |
+
"learning_rate": 4.879e-05,
|
| 704 |
+
"lorentz_dist_loss": 0.8342,
|
| 705 |
+
"loss": 10.05126495361328,
|
| 706 |
+
"step": 870
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.01418468296024662,
|
| 710 |
+
"grad_norm": 511.6054382324219,
|
| 711 |
+
"learning_rate": 4.935666666666667e-05,
|
| 712 |
+
"lorentz_dist_loss": 0.8345,
|
| 713 |
+
"loss": 10.05462646484375,
|
| 714 |
+
"step": 880
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.014345872539340332,
|
| 718 |
+
"grad_norm": 441.66339111328125,
|
| 719 |
+
"learning_rate": 4.992333333333334e-05,
|
| 720 |
+
"lorentz_dist_loss": 0.8331,
|
| 721 |
+
"loss": 10.038479614257813,
|
| 722 |
+
"step": 890
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.014507062118434043,
|
| 726 |
+
"grad_norm": 517.2638549804688,
|
| 727 |
+
"learning_rate": 5.049e-05,
|
| 728 |
+
"lorentz_dist_loss": 0.8318,
|
| 729 |
+
"loss": 10.022047424316407,
|
| 730 |
+
"step": 900
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.014668251697527755,
|
| 734 |
+
"grad_norm": 484.39080810546875,
|
| 735 |
+
"learning_rate": 5.105666666666667e-05,
|
| 736 |
+
"lorentz_dist_loss": 0.8332,
|
| 737 |
+
"loss": 10.039347076416016,
|
| 738 |
+
"step": 910
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.014829441276621466,
|
| 742 |
+
"grad_norm": 452.8804016113281,
|
| 743 |
+
"learning_rate": 5.162333333333333e-05,
|
| 744 |
+
"lorentz_dist_loss": 0.832,
|
| 745 |
+
"loss": 10.025006103515626,
|
| 746 |
+
"step": 920
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.014990630855715179,
|
| 750 |
+
"grad_norm": 454.01776123046875,
|
| 751 |
+
"learning_rate": 5.219e-05,
|
| 752 |
+
"lorentz_dist_loss": 0.8291,
|
| 753 |
+
"loss": 9.98986587524414,
|
| 754 |
+
"step": 930
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.01515182043480889,
|
| 758 |
+
"grad_norm": 495.3049011230469,
|
| 759 |
+
"learning_rate": 5.2756666666666676e-05,
|
| 760 |
+
"lorentz_dist_loss": 0.8293,
|
| 761 |
+
"loss": 9.992280578613281,
|
| 762 |
+
"step": 940
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.015313010013902601,
|
| 766 |
+
"grad_norm": 479.06365966796875,
|
| 767 |
+
"learning_rate": 5.332333333333333e-05,
|
| 768 |
+
"lorentz_dist_loss": 0.8306,
|
| 769 |
+
"loss": 10.008704376220702,
|
| 770 |
+
"step": 950
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.015474199592996313,
|
| 774 |
+
"grad_norm": 435.234375,
|
| 775 |
+
"learning_rate": 5.389000000000001e-05,
|
| 776 |
+
"lorentz_dist_loss": 0.8303,
|
| 777 |
+
"loss": 10.004000854492187,
|
| 778 |
+
"step": 960
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 0.015635389172090026,
|
| 782 |
+
"grad_norm": 391.1128234863281,
|
| 783 |
+
"learning_rate": 5.445666666666668e-05,
|
| 784 |
+
"lorentz_dist_loss": 0.8302,
|
| 785 |
+
"loss": 10.003079986572265,
|
| 786 |
+
"step": 970
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.015796578751183735,
|
| 790 |
+
"grad_norm": 434.96502685546875,
|
| 791 |
+
"learning_rate": 5.502333333333334e-05,
|
| 792 |
+
"lorentz_dist_loss": 0.8278,
|
| 793 |
+
"loss": 9.974259185791016,
|
| 794 |
+
"step": 980
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.01595776833027745,
|
| 798 |
+
"grad_norm": 370.57757568359375,
|
| 799 |
+
"learning_rate": 5.559000000000001e-05,
|
| 800 |
+
"lorentz_dist_loss": 0.8268,
|
| 801 |
+
"loss": 9.96210174560547,
|
| 802 |
+
"step": 990
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.016118957909371158,
|
| 806 |
+
"grad_norm": 516.6535034179688,
|
| 807 |
+
"learning_rate": 5.6156666666666664e-05,
|
| 808 |
+
"lorentz_dist_loss": 0.8276,
|
| 809 |
+
"loss": 9.97235565185547,
|
| 810 |
+
"step": 1000
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.01628014748846487,
|
| 814 |
+
"grad_norm": 1087.721923828125,
|
| 815 |
+
"learning_rate": 5.672333333333334e-05,
|
| 816 |
+
"lorentz_dist_loss": 0.8265,
|
| 817 |
+
"loss": 9.959116363525391,
|
| 818 |
+
"step": 1010
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.016441337067558584,
|
| 822 |
+
"grad_norm": 538.1283569335938,
|
| 823 |
+
"learning_rate": 5.729000000000001e-05,
|
| 824 |
+
"lorentz_dist_loss": 0.8267,
|
| 825 |
+
"loss": 9.961454010009765,
|
| 826 |
+
"step": 1020
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.016602526646652294,
|
| 830 |
+
"grad_norm": 411.9395446777344,
|
| 831 |
+
"learning_rate": 5.785666666666667e-05,
|
| 832 |
+
"lorentz_dist_loss": 0.8257,
|
| 833 |
+
"loss": 9.948921203613281,
|
| 834 |
+
"step": 1030
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 0.016763716225746007,
|
| 838 |
+
"grad_norm": 443.6706237792969,
|
| 839 |
+
"learning_rate": 5.842333333333334e-05,
|
| 840 |
+
"lorentz_dist_loss": 0.8241,
|
| 841 |
+
"loss": 9.930237579345704,
|
| 842 |
+
"step": 1040
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.016924905804839716,
|
| 846 |
+
"grad_norm": 416.83306884765625,
|
| 847 |
+
"learning_rate": 5.899e-05,
|
| 848 |
+
"lorentz_dist_loss": 0.8239,
|
| 849 |
+
"loss": 9.92744140625,
|
| 850 |
+
"step": 1050
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.01708609538393343,
|
| 854 |
+
"grad_norm": 376.559326171875,
|
| 855 |
+
"learning_rate": 5.955666666666667e-05,
|
| 856 |
+
"lorentz_dist_loss": 0.8244,
|
| 857 |
+
"loss": 9.933025360107422,
|
| 858 |
+
"step": 1060
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.01724728496302714,
|
| 862 |
+
"grad_norm": 469.279541015625,
|
| 863 |
+
"learning_rate": 6.012333333333334e-05,
|
| 864 |
+
"lorentz_dist_loss": 0.8236,
|
| 865 |
+
"loss": 9.92404556274414,
|
| 866 |
+
"step": 1070
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.017408474542120852,
|
| 870 |
+
"grad_norm": 477.0194396972656,
|
| 871 |
+
"learning_rate": 6.069e-05,
|
| 872 |
+
"lorentz_dist_loss": 0.8248,
|
| 873 |
+
"loss": 9.93814468383789,
|
| 874 |
+
"step": 1080
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.017569664121214565,
|
| 878 |
+
"grad_norm": 406.7455139160156,
|
| 879 |
+
"learning_rate": 6.125666666666667e-05,
|
| 880 |
+
"lorentz_dist_loss": 0.8217,
|
| 881 |
+
"loss": 9.901485443115234,
|
| 882 |
+
"step": 1090
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.017730853700308274,
|
| 886 |
+
"grad_norm": 634.7399291992188,
|
| 887 |
+
"learning_rate": 6.182333333333334e-05,
|
| 888 |
+
"lorentz_dist_loss": 0.8211,
|
| 889 |
+
"loss": 9.894142150878906,
|
| 890 |
+
"step": 1100
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.017892043279401987,
|
| 894 |
+
"grad_norm": 339.6075134277344,
|
| 895 |
+
"learning_rate": 6.239e-05,
|
| 896 |
+
"lorentz_dist_loss": 0.8213,
|
| 897 |
+
"loss": 9.896692657470703,
|
| 898 |
+
"step": 1110
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.018053232858495697,
|
| 902 |
+
"grad_norm": 440.0519104003906,
|
| 903 |
+
"learning_rate": 6.295666666666668e-05,
|
| 904 |
+
"lorentz_dist_loss": 0.8177,
|
| 905 |
+
"loss": 9.852577209472656,
|
| 906 |
+
"step": 1120
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.01821442243758941,
|
| 910 |
+
"grad_norm": 413.9136047363281,
|
| 911 |
+
"learning_rate": 6.352333333333333e-05,
|
| 912 |
+
"lorentz_dist_loss": 0.8181,
|
| 913 |
+
"loss": 9.857941436767579,
|
| 914 |
+
"step": 1130
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.018375612016683123,
|
| 918 |
+
"grad_norm": 413.617919921875,
|
| 919 |
+
"learning_rate": 6.409e-05,
|
| 920 |
+
"lorentz_dist_loss": 0.819,
|
| 921 |
+
"loss": 9.868207550048828,
|
| 922 |
+
"step": 1140
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.018536801595776833,
|
| 926 |
+
"grad_norm": 594.518310546875,
|
| 927 |
+
"learning_rate": 6.465666666666668e-05,
|
| 928 |
+
"lorentz_dist_loss": 0.8169,
|
| 929 |
+
"loss": 9.84332504272461,
|
| 930 |
+
"step": 1150
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.018697991174870546,
|
| 934 |
+
"grad_norm": 357.36077880859375,
|
| 935 |
+
"learning_rate": 6.522333333333333e-05,
|
| 936 |
+
"lorentz_dist_loss": 0.8152,
|
| 937 |
+
"loss": 9.82244873046875,
|
| 938 |
+
"step": 1160
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.018859180753964255,
|
| 942 |
+
"grad_norm": 405.4687194824219,
|
| 943 |
+
"learning_rate": 6.579e-05,
|
| 944 |
+
"lorentz_dist_loss": 0.8153,
|
| 945 |
+
"loss": 9.824163055419922,
|
| 946 |
+
"step": 1170
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.01902037033305797,
|
| 950 |
+
"grad_norm": 358.0759582519531,
|
| 951 |
+
"learning_rate": 6.635666666666667e-05,
|
| 952 |
+
"lorentz_dist_loss": 0.8156,
|
| 953 |
+
"loss": 9.827672576904297,
|
| 954 |
+
"step": 1180
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.019181559912151678,
|
| 958 |
+
"grad_norm": 336.9348449707031,
|
| 959 |
+
"learning_rate": 6.692333333333334e-05,
|
| 960 |
+
"lorentz_dist_loss": 0.8149,
|
| 961 |
+
"loss": 9.8189453125,
|
| 962 |
+
"step": 1190
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.01934274949124539,
|
| 966 |
+
"grad_norm": 364.91217041015625,
|
| 967 |
+
"learning_rate": 6.749e-05,
|
| 968 |
+
"lorentz_dist_loss": 0.8139,
|
| 969 |
+
"loss": 9.807315826416016,
|
| 970 |
+
"step": 1200
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.019503939070339104,
|
| 974 |
+
"grad_norm": 621.1389770507812,
|
| 975 |
+
"learning_rate": 6.805666666666667e-05,
|
| 976 |
+
"lorentz_dist_loss": 0.8131,
|
| 977 |
+
"loss": 9.797688293457032,
|
| 978 |
+
"step": 1210
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.019665128649432814,
|
| 982 |
+
"grad_norm": 556.9744873046875,
|
| 983 |
+
"learning_rate": 6.862333333333334e-05,
|
| 984 |
+
"lorentz_dist_loss": 0.8137,
|
| 985 |
+
"loss": 9.805026245117187,
|
| 986 |
+
"step": 1220
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.019826318228526527,
|
| 990 |
+
"grad_norm": 367.6215515136719,
|
| 991 |
+
"learning_rate": 6.919e-05,
|
| 992 |
+
"lorentz_dist_loss": 0.8117,
|
| 993 |
+
"loss": 9.78070068359375,
|
| 994 |
+
"step": 1230
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.019987507807620236,
|
| 998 |
+
"grad_norm": 538.5054931640625,
|
| 999 |
+
"learning_rate": 6.975666666666667e-05,
|
| 1000 |
+
"lorentz_dist_loss": 0.8119,
|
| 1001 |
+
"loss": 9.782686614990235,
|
| 1002 |
+
"step": 1240
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.02014869738671395,
|
| 1006 |
+
"grad_norm": 541.1419067382812,
|
| 1007 |
+
"learning_rate": 7.032333333333334e-05,
|
| 1008 |
+
"lorentz_dist_loss": 0.8104,
|
| 1009 |
+
"loss": 9.764971923828124,
|
| 1010 |
+
"step": 1250
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.020309886965807662,
|
| 1014 |
+
"grad_norm": 745.2047729492188,
|
| 1015 |
+
"learning_rate": 7.089000000000001e-05,
|
| 1016 |
+
"lorentz_dist_loss": 0.8087,
|
| 1017 |
+
"loss": 9.745184326171875,
|
| 1018 |
+
"step": 1260
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.020471076544901372,
|
| 1022 |
+
"grad_norm": 438.1859130859375,
|
| 1023 |
+
"learning_rate": 7.145666666666667e-05,
|
| 1024 |
+
"lorentz_dist_loss": 0.8099,
|
| 1025 |
+
"loss": 9.759564208984376,
|
| 1026 |
+
"step": 1270
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.020632266123995085,
|
| 1030 |
+
"grad_norm": 428.1754455566406,
|
| 1031 |
+
"learning_rate": 7.202333333333335e-05,
|
| 1032 |
+
"lorentz_dist_loss": 0.8095,
|
| 1033 |
+
"loss": 9.753829956054688,
|
| 1034 |
+
"step": 1280
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.020793455703088794,
|
| 1038 |
+
"grad_norm": 411.3019104003906,
|
| 1039 |
+
"learning_rate": 7.259000000000001e-05,
|
| 1040 |
+
"lorentz_dist_loss": 0.8062,
|
| 1041 |
+
"loss": 9.714922332763672,
|
| 1042 |
+
"step": 1290
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 0.020954645282182507,
|
| 1046 |
+
"grad_norm": 473.98443603515625,
|
| 1047 |
+
"learning_rate": 7.315666666666667e-05,
|
| 1048 |
+
"lorentz_dist_loss": 0.8075,
|
| 1049 |
+
"loss": 9.730874633789062,
|
| 1050 |
+
"step": 1300
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.021115834861276217,
|
| 1054 |
+
"grad_norm": 810.3408813476562,
|
| 1055 |
+
"learning_rate": 7.372333333333333e-05,
|
| 1056 |
+
"lorentz_dist_loss": 0.807,
|
| 1057 |
+
"loss": 9.724625396728516,
|
| 1058 |
+
"step": 1310
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.02127702444036993,
|
| 1062 |
+
"grad_norm": 587.3361206054688,
|
| 1063 |
+
"learning_rate": 7.429000000000001e-05,
|
| 1064 |
+
"lorentz_dist_loss": 0.8064,
|
| 1065 |
+
"loss": 9.717549133300782,
|
| 1066 |
+
"step": 1320
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.021438214019463643,
|
| 1070 |
+
"grad_norm": 304.0115661621094,
|
| 1071 |
+
"learning_rate": 7.485666666666667e-05,
|
| 1072 |
+
"lorentz_dist_loss": 0.8043,
|
| 1073 |
+
"loss": 9.691986083984375,
|
| 1074 |
+
"step": 1330
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.021599403598557353,
|
| 1078 |
+
"grad_norm": 361.18267822265625,
|
| 1079 |
+
"learning_rate": 7.542333333333333e-05,
|
| 1080 |
+
"lorentz_dist_loss": 0.806,
|
| 1081 |
+
"loss": 9.711846160888673,
|
| 1082 |
+
"step": 1340
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.021760593177651066,
|
| 1086 |
+
"grad_norm": 748.0859985351562,
|
| 1087 |
+
"learning_rate": 7.599000000000001e-05,
|
| 1088 |
+
"lorentz_dist_loss": 0.8033,
|
| 1089 |
+
"loss": 9.679519653320312,
|
| 1090 |
+
"step": 1350
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.021921782756744775,
|
| 1094 |
+
"grad_norm": 859.8388061523438,
|
| 1095 |
+
"learning_rate": 7.655666666666667e-05,
|
| 1096 |
+
"lorentz_dist_loss": 0.806,
|
| 1097 |
+
"loss": 9.712540435791016,
|
| 1098 |
+
"step": 1360
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.02208297233583849,
|
| 1102 |
+
"grad_norm": 378.6498107910156,
|
| 1103 |
+
"learning_rate": 7.712333333333333e-05,
|
| 1104 |
+
"lorentz_dist_loss": 0.8032,
|
| 1105 |
+
"loss": 9.679087066650391,
|
| 1106 |
+
"step": 1370
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.0222441619149322,
|
| 1110 |
+
"grad_norm": 1007.4165649414062,
|
| 1111 |
+
"learning_rate": 7.769000000000001e-05,
|
| 1112 |
+
"lorentz_dist_loss": 0.8017,
|
| 1113 |
+
"loss": 9.660994720458984,
|
| 1114 |
+
"step": 1380
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 0.02240535149402591,
|
| 1118 |
+
"grad_norm": 480.6250305175781,
|
| 1119 |
+
"learning_rate": 7.825666666666667e-05,
|
| 1120 |
+
"lorentz_dist_loss": 0.8005,
|
| 1121 |
+
"loss": 9.645849609375,
|
| 1122 |
+
"step": 1390
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.022566541073119624,
|
| 1126 |
+
"grad_norm": 361.4166259765625,
|
| 1127 |
+
"learning_rate": 7.882333333333333e-05,
|
| 1128 |
+
"lorentz_dist_loss": 0.8023,
|
| 1129 |
+
"loss": 9.667982482910157,
|
| 1130 |
+
"step": 1400
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.022727730652213333,
|
| 1134 |
+
"grad_norm": 453.7378234863281,
|
| 1135 |
+
"learning_rate": 7.939000000000001e-05,
|
| 1136 |
+
"lorentz_dist_loss": 0.801,
|
| 1137 |
+
"loss": 9.652360534667968,
|
| 1138 |
+
"step": 1410
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.022888920231307047,
|
| 1142 |
+
"grad_norm": 383.8326110839844,
|
| 1143 |
+
"learning_rate": 7.995666666666667e-05,
|
| 1144 |
+
"lorentz_dist_loss": 0.8008,
|
| 1145 |
+
"loss": 9.65002670288086,
|
| 1146 |
+
"step": 1420
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.023050109810400756,
|
| 1150 |
+
"grad_norm": 321.88232421875,
|
| 1151 |
+
"learning_rate": 8.052333333333334e-05,
|
| 1152 |
+
"lorentz_dist_loss": 0.8012,
|
| 1153 |
+
"loss": 9.653956604003906,
|
| 1154 |
+
"step": 1430
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.02321129938949447,
|
| 1158 |
+
"grad_norm": 318.3651428222656,
|
| 1159 |
+
"learning_rate": 8.109e-05,
|
| 1160 |
+
"lorentz_dist_loss": 0.7988,
|
| 1161 |
+
"loss": 9.625229644775391,
|
| 1162 |
+
"step": 1440
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 0.023372488968588182,
|
| 1166 |
+
"grad_norm": 319.4456481933594,
|
| 1167 |
+
"learning_rate": 8.165666666666667e-05,
|
| 1168 |
+
"lorentz_dist_loss": 0.7977,
|
| 1169 |
+
"loss": 9.61263427734375,
|
| 1170 |
+
"step": 1450
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 0.023533678547681892,
|
| 1174 |
+
"grad_norm": 839.595703125,
|
| 1175 |
+
"learning_rate": 8.222333333333335e-05,
|
| 1176 |
+
"lorentz_dist_loss": 0.798,
|
| 1177 |
+
"loss": 9.616504669189453,
|
| 1178 |
+
"step": 1460
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.023694868126775605,
|
| 1182 |
+
"grad_norm": 299.8286437988281,
|
| 1183 |
+
"learning_rate": 8.279e-05,
|
| 1184 |
+
"lorentz_dist_loss": 0.7957,
|
| 1185 |
+
"loss": 9.58837432861328,
|
| 1186 |
+
"step": 1470
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.023856057705869314,
|
| 1190 |
+
"grad_norm": 359.0120849609375,
|
| 1191 |
+
"learning_rate": 8.335666666666667e-05,
|
| 1192 |
+
"lorentz_dist_loss": 0.7955,
|
| 1193 |
+
"loss": 9.585936737060546,
|
| 1194 |
+
"step": 1480
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.024017247284963027,
|
| 1198 |
+
"grad_norm": 345.22625732421875,
|
| 1199 |
+
"learning_rate": 8.392333333333334e-05,
|
| 1200 |
+
"lorentz_dist_loss": 0.7945,
|
| 1201 |
+
"loss": 9.57409896850586,
|
| 1202 |
+
"step": 1490
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.02417843686405674,
|
| 1206 |
+
"grad_norm": 407.0867919921875,
|
| 1207 |
+
"learning_rate": 8.449e-05,
|
| 1208 |
+
"lorentz_dist_loss": 0.7937,
|
| 1209 |
+
"loss": 9.564083099365234,
|
| 1210 |
+
"step": 1500
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.02433962644315045,
|
| 1214 |
+
"grad_norm": 349.28125,
|
| 1215 |
+
"learning_rate": 8.499999994277469e-05,
|
| 1216 |
+
"lorentz_dist_loss": 0.7924,
|
| 1217 |
+
"loss": 9.549102783203125,
|
| 1218 |
+
"step": 1510
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 0.024500816022244163,
|
| 1222 |
+
"grad_norm": 599.9000854492188,
|
| 1223 |
+
"learning_rate": 8.499999307573704e-05,
|
| 1224 |
+
"lorentz_dist_loss": 0.7913,
|
| 1225 |
+
"loss": 9.534966278076173,
|
| 1226 |
+
"step": 1520
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.024662005601337873,
|
| 1230 |
+
"grad_norm": 226.55606079101562,
|
| 1231 |
+
"learning_rate": 8.499997476363848e-05,
|
| 1232 |
+
"lorentz_dist_loss": 0.7923,
|
| 1233 |
+
"loss": 9.547154235839844,
|
| 1234 |
+
"step": 1530
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.024823195180431586,
|
| 1238 |
+
"grad_norm": 262.4807434082031,
|
| 1239 |
+
"learning_rate": 8.499994500648389e-05,
|
| 1240 |
+
"lorentz_dist_loss": 0.7911,
|
| 1241 |
+
"loss": 9.533607482910156,
|
| 1242 |
+
"step": 1540
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.024984384759525295,
|
| 1246 |
+
"grad_norm": 255.0460968017578,
|
| 1247 |
+
"learning_rate": 8.499990380428132e-05,
|
| 1248 |
+
"lorentz_dist_loss": 0.7914,
|
| 1249 |
+
"loss": 9.536710357666015,
|
| 1250 |
+
"step": 1550
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.025145574338619008,
|
| 1254 |
+
"grad_norm": 202.52281188964844,
|
| 1255 |
+
"learning_rate": 8.499985115704187e-05,
|
| 1256 |
+
"lorentz_dist_loss": 0.789,
|
| 1257 |
+
"loss": 9.507820892333985,
|
| 1258 |
+
"step": 1560
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.02530676391771272,
|
| 1262 |
+
"grad_norm": 187.66456604003906,
|
| 1263 |
+
"learning_rate": 8.499978706477969e-05,
|
| 1264 |
+
"lorentz_dist_loss": 0.7883,
|
| 1265 |
+
"loss": 9.4996826171875,
|
| 1266 |
+
"step": 1570
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.02546795349680643,
|
| 1270 |
+
"grad_norm": 153.35372924804688,
|
| 1271 |
+
"learning_rate": 8.499971152751206e-05,
|
| 1272 |
+
"lorentz_dist_loss": 0.7858,
|
| 1273 |
+
"loss": 9.46889877319336,
|
| 1274 |
+
"step": 1580
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.025629143075900144,
|
| 1278 |
+
"grad_norm": 470.9952697753906,
|
| 1279 |
+
"learning_rate": 8.499962454525932e-05,
|
| 1280 |
+
"lorentz_dist_loss": 0.7865,
|
| 1281 |
+
"loss": 9.478172302246094,
|
| 1282 |
+
"step": 1590
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 0.025790332654993853,
|
| 1286 |
+
"grad_norm": 340.3394775390625,
|
| 1287 |
+
"learning_rate": 8.49995261180449e-05,
|
| 1288 |
+
"lorentz_dist_loss": 0.7853,
|
| 1289 |
+
"loss": 9.463638305664062,
|
| 1290 |
+
"step": 1600
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.025951522234087566,
|
| 1294 |
+
"grad_norm": 268.3125305175781,
|
| 1295 |
+
"learning_rate": 8.499941624589527e-05,
|
| 1296 |
+
"lorentz_dist_loss": 0.7857,
|
| 1297 |
+
"loss": 9.467852783203124,
|
| 1298 |
+
"step": 1610
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.02611271181318128,
|
| 1302 |
+
"grad_norm": 290.41497802734375,
|
| 1303 |
+
"learning_rate": 8.499929492884008e-05,
|
| 1304 |
+
"lorentz_dist_loss": 0.7848,
|
| 1305 |
+
"loss": 9.457717895507812,
|
| 1306 |
+
"step": 1620
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.02627390139227499,
|
| 1310 |
+
"grad_norm": 133.57717895507812,
|
| 1311 |
+
"learning_rate": 8.499916216691193e-05,
|
| 1312 |
+
"lorentz_dist_loss": 0.7833,
|
| 1313 |
+
"loss": 9.438694763183594,
|
| 1314 |
+
"step": 1630
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.026435090971368702,
|
| 1318 |
+
"grad_norm": 278.4210205078125,
|
| 1319 |
+
"learning_rate": 8.499901796014664e-05,
|
| 1320 |
+
"lorentz_dist_loss": 0.7846,
|
| 1321 |
+
"loss": 9.454314422607421,
|
| 1322 |
+
"step": 1640
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.02659628055046241,
|
| 1326 |
+
"grad_norm": 109.79666137695312,
|
| 1327 |
+
"learning_rate": 8.499886230858299e-05,
|
| 1328 |
+
"lorentz_dist_loss": 0.7818,
|
| 1329 |
+
"loss": 9.421340942382812,
|
| 1330 |
+
"step": 1650
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 0.026757470129556125,
|
| 1334 |
+
"grad_norm": 89.1125717163086,
|
| 1335 |
+
"learning_rate": 8.499869521226292e-05,
|
| 1336 |
+
"lorentz_dist_loss": 0.7791,
|
| 1337 |
+
"loss": 9.38824234008789,
|
| 1338 |
+
"step": 1660
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.026918659708649834,
|
| 1342 |
+
"grad_norm": 274.308837890625,
|
| 1343 |
+
"learning_rate": 8.499851667123144e-05,
|
| 1344 |
+
"lorentz_dist_loss": 0.7797,
|
| 1345 |
+
"loss": 9.395606231689452,
|
| 1346 |
+
"step": 1670
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.027079849287743547,
|
| 1350 |
+
"grad_norm": 469.7413330078125,
|
| 1351 |
+
"learning_rate": 8.49983266855366e-05,
|
| 1352 |
+
"lorentz_dist_loss": 0.7804,
|
| 1353 |
+
"loss": 9.404562377929688,
|
| 1354 |
+
"step": 1680
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.02724103886683726,
|
| 1358 |
+
"grad_norm": 149.0135955810547,
|
| 1359 |
+
"learning_rate": 8.499812525522958e-05,
|
| 1360 |
+
"lorentz_dist_loss": 0.7816,
|
| 1361 |
+
"loss": 9.418919372558594,
|
| 1362 |
+
"step": 1690
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 0.02740222844593097,
|
| 1366 |
+
"grad_norm": 110.48833465576172,
|
| 1367 |
+
"learning_rate": 8.499791238036465e-05,
|
| 1368 |
+
"lorentz_dist_loss": 0.7785,
|
| 1369 |
+
"loss": 9.38119888305664,
|
| 1370 |
+
"step": 1700
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.027563418025024683,
|
| 1374 |
+
"grad_norm": 117.34359741210938,
|
| 1375 |
+
"learning_rate": 8.499768806099907e-05,
|
| 1376 |
+
"lorentz_dist_loss": 0.7794,
|
| 1377 |
+
"loss": 9.392098236083985,
|
| 1378 |
+
"step": 1710
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.027724607604118393,
|
| 1382 |
+
"grad_norm": 77.1554183959961,
|
| 1383 |
+
"learning_rate": 8.499745229719331e-05,
|
| 1384 |
+
"lorentz_dist_loss": 0.7774,
|
| 1385 |
+
"loss": 9.368196105957031,
|
| 1386 |
+
"step": 1720
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 0.027885797183212106,
|
| 1390 |
+
"grad_norm": 146.49072265625,
|
| 1391 |
+
"learning_rate": 8.499720508901084e-05,
|
| 1392 |
+
"lorentz_dist_loss": 0.7766,
|
| 1393 |
+
"loss": 9.358159637451172,
|
| 1394 |
+
"step": 1730
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.02804698676230582,
|
| 1398 |
+
"grad_norm": 105.40322875976562,
|
| 1399 |
+
"learning_rate": 8.499694643651821e-05,
|
| 1400 |
+
"lorentz_dist_loss": 0.7756,
|
| 1401 |
+
"loss": 9.346617126464844,
|
| 1402 |
+
"step": 1740
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.028208176341399528,
|
| 1406 |
+
"grad_norm": 74.9076919555664,
|
| 1407 |
+
"learning_rate": 8.499667633978512e-05,
|
| 1408 |
+
"lorentz_dist_loss": 0.7743,
|
| 1409 |
+
"loss": 9.331331634521485,
|
| 1410 |
+
"step": 1750
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.02836936592049324,
|
| 1414 |
+
"grad_norm": 103.18841552734375,
|
| 1415 |
+
"learning_rate": 8.499639479888426e-05,
|
| 1416 |
+
"lorentz_dist_loss": 0.7747,
|
| 1417 |
+
"loss": 9.335555267333984,
|
| 1418 |
+
"step": 1760
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.02853055549958695,
|
| 1422 |
+
"grad_norm": 74.23155212402344,
|
| 1423 |
+
"learning_rate": 8.499610181389148e-05,
|
| 1424 |
+
"lorentz_dist_loss": 0.7727,
|
| 1425 |
+
"loss": 9.311741638183594,
|
| 1426 |
+
"step": 1770
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.028691745078680664,
|
| 1430 |
+
"grad_norm": 84.15802764892578,
|
| 1431 |
+
"learning_rate": 8.499579738488565e-05,
|
| 1432 |
+
"lorentz_dist_loss": 0.7718,
|
| 1433 |
+
"loss": 9.300537109375,
|
| 1434 |
+
"step": 1780
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.028852934657774373,
|
| 1438 |
+
"grad_norm": 80.7850570678711,
|
| 1439 |
+
"learning_rate": 8.499548151194878e-05,
|
| 1440 |
+
"lorentz_dist_loss": 0.7703,
|
| 1441 |
+
"loss": 9.282907104492187,
|
| 1442 |
+
"step": 1790
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.029014124236868086,
|
| 1446 |
+
"grad_norm": 95.6905288696289,
|
| 1447 |
+
"learning_rate": 8.499515419516594e-05,
|
| 1448 |
+
"lorentz_dist_loss": 0.7699,
|
| 1449 |
+
"loss": 9.277885437011719,
|
| 1450 |
+
"step": 1800
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 0.0291753138159618,
|
| 1454 |
+
"grad_norm": 4876.93896484375,
|
| 1455 |
+
"learning_rate": 8.499481543462524e-05,
|
| 1456 |
+
"lorentz_dist_loss": 1.0296,
|
| 1457 |
+
"loss": 12.396101379394532,
|
| 1458 |
+
"step": 1810
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.02933650339505551,
|
| 1462 |
+
"grad_norm": 4075.185791015625,
|
| 1463 |
+
"learning_rate": 8.49944652304179e-05,
|
| 1464 |
+
"lorentz_dist_loss": 0.9462,
|
| 1465 |
+
"loss": 11.395193481445313,
|
| 1466 |
+
"step": 1820
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.029497692974149222,
|
| 1470 |
+
"grad_norm": 2967.10107421875,
|
| 1471 |
+
"learning_rate": 8.499410358263828e-05,
|
| 1472 |
+
"lorentz_dist_loss": 0.9254,
|
| 1473 |
+
"loss": 11.146472930908203,
|
| 1474 |
+
"step": 1830
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 0.02965888255324293,
|
| 1478 |
+
"grad_norm": 2984.347412109375,
|
| 1479 |
+
"learning_rate": 8.499373049138374e-05,
|
| 1480 |
+
"lorentz_dist_loss": 0.9147,
|
| 1481 |
+
"loss": 11.018094635009765,
|
| 1482 |
+
"step": 1840
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.029820072132336645,
|
| 1486 |
+
"grad_norm": 2961.19091796875,
|
| 1487 |
+
"learning_rate": 8.499334595675474e-05,
|
| 1488 |
+
"lorentz_dist_loss": 0.9057,
|
| 1489 |
+
"loss": 10.90991439819336,
|
| 1490 |
+
"step": 1850
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.029981261711430358,
|
| 1494 |
+
"grad_norm": 2434.526611328125,
|
| 1495 |
+
"learning_rate": 8.499294997885486e-05,
|
| 1496 |
+
"lorentz_dist_loss": 0.9013,
|
| 1497 |
+
"loss": 10.856851196289062,
|
| 1498 |
+
"step": 1860
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 0.030142451290524067,
|
| 1502 |
+
"grad_norm": 2446.460693359375,
|
| 1503 |
+
"learning_rate": 8.499254255779072e-05,
|
| 1504 |
+
"lorentz_dist_loss": 0.8955,
|
| 1505 |
+
"loss": 10.787834167480469,
|
| 1506 |
+
"step": 1870
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.03030364086961778,
|
| 1510 |
+
"grad_norm": 2283.36572265625,
|
| 1511 |
+
"learning_rate": 8.499212369367203e-05,
|
| 1512 |
+
"lorentz_dist_loss": 0.8917,
|
| 1513 |
+
"loss": 10.741209411621094,
|
| 1514 |
+
"step": 1880
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.03046483044871149,
|
| 1518 |
+
"grad_norm": 2228.294189453125,
|
| 1519 |
+
"learning_rate": 8.499169338661161e-05,
|
| 1520 |
+
"lorentz_dist_loss": 0.8885,
|
| 1521 |
+
"loss": 10.70361557006836,
|
| 1522 |
+
"step": 1890
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.030626020027805203,
|
| 1526 |
+
"grad_norm": 1952.3082275390625,
|
| 1527 |
+
"learning_rate": 8.499125163672533e-05,
|
| 1528 |
+
"lorentz_dist_loss": 0.8847,
|
| 1529 |
+
"loss": 10.657195281982421,
|
| 1530 |
+
"step": 1900
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.030787209606898912,
|
| 1534 |
+
"grad_norm": 1867.9254150390625,
|
| 1535 |
+
"learning_rate": 8.499079844413214e-05,
|
| 1536 |
+
"lorentz_dist_loss": 0.8814,
|
| 1537 |
+
"loss": 10.617882537841798,
|
| 1538 |
+
"step": 1910
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.030948399185992626,
|
| 1542 |
+
"grad_norm": 1723.9246826171875,
|
| 1543 |
+
"learning_rate": 8.499033380895408e-05,
|
| 1544 |
+
"lorentz_dist_loss": 0.8782,
|
| 1545 |
+
"loss": 10.579761505126953,
|
| 1546 |
+
"step": 1920
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.03110958876508634,
|
| 1550 |
+
"grad_norm": 1717.7218017578125,
|
| 1551 |
+
"learning_rate": 8.49898577313163e-05,
|
| 1552 |
+
"lorentz_dist_loss": 0.8759,
|
| 1553 |
+
"loss": 10.552434539794922,
|
| 1554 |
+
"step": 1930
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.03127077834418005,
|
| 1558 |
+
"grad_norm": 1879.3978271484375,
|
| 1559 |
+
"learning_rate": 8.498937021134698e-05,
|
| 1560 |
+
"lorentz_dist_loss": 0.8753,
|
| 1561 |
+
"loss": 10.54498062133789,
|
| 1562 |
+
"step": 1940
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.03143196792327376,
|
| 1566 |
+
"grad_norm": 1633.0048828125,
|
| 1567 |
+
"learning_rate": 8.498887124917744e-05,
|
| 1568 |
+
"lorentz_dist_loss": 0.873,
|
| 1569 |
+
"loss": 10.517025756835938,
|
| 1570 |
+
"step": 1950
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.03159315750236747,
|
| 1574 |
+
"grad_norm": 2058.17333984375,
|
| 1575 |
+
"learning_rate": 8.4988360844942e-05,
|
| 1576 |
+
"lorentz_dist_loss": 0.8709,
|
| 1577 |
+
"loss": 10.492323303222657,
|
| 1578 |
+
"step": 1960
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.03175434708146118,
|
| 1582 |
+
"grad_norm": 1560.37890625,
|
| 1583 |
+
"learning_rate": 8.498783899877814e-05,
|
| 1584 |
+
"lorentz_dist_loss": 0.8698,
|
| 1585 |
+
"loss": 10.478605651855469,
|
| 1586 |
+
"step": 1970
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.0319155366605549,
|
| 1590 |
+
"grad_norm": 1542.8436279296875,
|
| 1591 |
+
"learning_rate": 8.49873057108264e-05,
|
| 1592 |
+
"lorentz_dist_loss": 0.8664,
|
| 1593 |
+
"loss": 10.438361358642577,
|
| 1594 |
+
"step": 1980
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.032076726239648606,
|
| 1598 |
+
"grad_norm": 1222.5784912109375,
|
| 1599 |
+
"learning_rate": 8.498676098123037e-05,
|
| 1600 |
+
"lorentz_dist_loss": 0.8648,
|
| 1601 |
+
"loss": 10.41856689453125,
|
| 1602 |
+
"step": 1990
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 0.032237915818742316,
|
| 1606 |
+
"grad_norm": 1264.2506103515625,
|
| 1607 |
+
"learning_rate": 8.498620481013676e-05,
|
| 1608 |
+
"lorentz_dist_loss": 0.8613,
|
| 1609 |
+
"loss": 10.377261352539062,
|
| 1610 |
+
"step": 2000
|
| 1611 |
+
}
|
| 1612 |
+
],
|
| 1613 |
+
"logging_steps": 10,
|
| 1614 |
+
"max_steps": 62039,
|
| 1615 |
+
"num_input_tokens_seen": 0,
|
| 1616 |
+
"num_train_epochs": 1,
|
| 1617 |
+
"save_steps": 200,
|
| 1618 |
+
"stateful_callbacks": {
|
| 1619 |
+
"TrainerControl": {
|
| 1620 |
+
"args": {
|
| 1621 |
+
"should_epoch_stop": false,
|
| 1622 |
+
"should_evaluate": false,
|
| 1623 |
+
"should_log": false,
|
| 1624 |
+
"should_save": true,
|
| 1625 |
+
"should_training_stop": false
|
| 1626 |
+
},
|
| 1627 |
+
"attributes": {}
|
| 1628 |
+
}
|
| 1629 |
+
},
|
| 1630 |
+
"total_flos": 3.71219262522624e+16,
|
| 1631 |
+
"train_batch_size": 12,
|
| 1632 |
+
"trial_name": null,
|
| 1633 |
+
"trial_params": null
|
| 1634 |
+
}
|
checkpoint-2000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24cb3627c0b189bac8b915bb2aba3a5fa64fb4b868b07c933223ab6deb5413e0
|
| 3 |
+
size 5201
|