Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- custom-tokenizer/full-finetuning/LLaMmlein_120M/config.json +30 -0
- custom-tokenizer/full-finetuning/LLaMmlein_120M/generation_config.json +6 -0
- custom-tokenizer/full-finetuning/LLaMmlein_120M/model.safetensors +3 -0
- custom-tokenizer/full-finetuning/LLaMmlein_120M/optimizer.pt +3 -0
- custom-tokenizer/full-finetuning/LLaMmlein_120M/rng_state.pth +3 -0
- custom-tokenizer/full-finetuning/LLaMmlein_120M/scheduler.pt +3 -0
- custom-tokenizer/full-finetuning/LLaMmlein_120M/trainer_state.json +1338 -0
- custom-tokenizer/full-finetuning/LLaMmlein_120M/training_args.bin +3 -0
- custom-tokenizer/full-finetuning/german-gpt2/config.json +40 -0
- custom-tokenizer/full-finetuning/german-gpt2/generation_config.json +6 -0
- custom-tokenizer/full-finetuning/german-gpt2/model.safetensors +3 -0
- custom-tokenizer/full-finetuning/german-gpt2/optimizer.pt +3 -0
- custom-tokenizer/full-finetuning/german-gpt2/rng_state.pth +3 -0
- custom-tokenizer/full-finetuning/german-gpt2/scheduler.pt +3 -0
- custom-tokenizer/full-finetuning/german-gpt2/trainer_state.json +1338 -0
- custom-tokenizer/full-finetuning/german-gpt2/training_args.bin +3 -0
- full-finetuning/LLaMmlein_120M/config.json +30 -0
- full-finetuning/LLaMmlein_120M/generation_config.json +6 -0
- full-finetuning/LLaMmlein_120M/model.safetensors +3 -0
- full-finetuning/LLaMmlein_120M/optimizer.pt +3 -0
- full-finetuning/LLaMmlein_120M/rng_state.pth +3 -0
- full-finetuning/LLaMmlein_120M/scheduler.pt +3 -0
- full-finetuning/LLaMmlein_120M/trainer_state.json +1338 -0
- full-finetuning/LLaMmlein_120M/training_args.bin +3 -0
- full-finetuning/german-gpt2/config.json +40 -0
- full-finetuning/german-gpt2/generation_config.json +6 -0
- full-finetuning/german-gpt2/model.safetensors +3 -0
- full-finetuning/german-gpt2/optimizer.pt +3 -0
- full-finetuning/german-gpt2/rng_state.pth +3 -0
- full-finetuning/german-gpt2/scheduler.pt +3 -0
- full-finetuning/german-gpt2/trainer_state.json +1338 -0
- full-finetuning/german-gpt2/training_args.bin +3 -0
- instruct-finetuning/base/LLaMmlein_120M/README.md +202 -0
- instruct-finetuning/base/LLaMmlein_120M/adapter_config.json +35 -0
- instruct-finetuning/base/LLaMmlein_120M/adapter_model.safetensors +3 -0
- instruct-finetuning/base/LLaMmlein_120M/optimizer.pt +3 -0
- instruct-finetuning/base/LLaMmlein_120M/rng_state.pth +3 -0
- instruct-finetuning/base/LLaMmlein_120M/scheduler.pt +3 -0
- instruct-finetuning/base/LLaMmlein_120M/special_tokens_map.json +24 -0
- instruct-finetuning/base/LLaMmlein_120M/tokenizer.json +0 -0
- instruct-finetuning/base/LLaMmlein_120M/tokenizer_config.json +43 -0
- instruct-finetuning/base/LLaMmlein_120M/trainer_state.json +1593 -0
- instruct-finetuning/base/LLaMmlein_120M/training_args.bin +3 -0
- instruct-finetuning/base/german-gpt2/README.md +202 -0
- instruct-finetuning/base/german-gpt2/adapter_config.json +35 -0
- instruct-finetuning/base/german-gpt2/adapter_model.safetensors +3 -0
- instruct-finetuning/base/german-gpt2/added_tokens.json +3 -0
- instruct-finetuning/base/german-gpt2/merges.txt +0 -0
- instruct-finetuning/base/german-gpt2/optimizer.pt +3 -0
- instruct-finetuning/base/german-gpt2/rng_state.pth +3 -0
custom-tokenizer/full-finetuning/LLaMmlein_120M/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "LSX-UniWue/LLaMmlein_120M",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 1,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"head_dim": 64,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 2048,
|
| 15 |
+
"max_position_embeddings": 2048,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"num_key_value_heads": 4,
|
| 21 |
+
"pretraining_tp": 1,
|
| 22 |
+
"rms_norm_eps": 1e-05,
|
| 23 |
+
"rope_scaling": null,
|
| 24 |
+
"rope_theta": 10000.0,
|
| 25 |
+
"tie_word_embeddings": false,
|
| 26 |
+
"torch_dtype": "float32",
|
| 27 |
+
"transformers_version": "4.47.1",
|
| 28 |
+
"use_cache": true,
|
| 29 |
+
"vocab_size": 32000
|
| 30 |
+
}
|
custom-tokenizer/full-finetuning/LLaMmlein_120M/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.47.1"
|
| 6 |
+
}
|
custom-tokenizer/full-finetuning/LLaMmlein_120M/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0b3044bcc64ac95bdd2658b8d0f9cabeb741f51fc1d99410a052b7d1c7182cb
|
| 3 |
+
size 498687008
|
custom-tokenizer/full-finetuning/LLaMmlein_120M/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08037fda9d57107296c2d1153683af7fc2effcc6947bfbe698b27c6809f44476
|
| 3 |
+
size 997443194
|
custom-tokenizer/full-finetuning/LLaMmlein_120M/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3bc1ac846859ad13fd356d3ed41fd20834205a7e8764e4198636863fca3f64c
|
| 3 |
+
size 14244
|
custom-tokenizer/full-finetuning/LLaMmlein_120M/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a4bab2c9813a0e45b4057cd5fbd9a5b26b99af9e15f94c8344ed33c80a85b8c
|
| 3 |
+
size 1064
|
custom-tokenizer/full-finetuning/LLaMmlein_120M/trainer_state.json
ADDED
|
@@ -0,0 +1,1338 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.157134532928467,
|
| 3 |
+
"best_model_checkpoint": "./models/custom-tokenizer/full-finetuning/LLaMmlein_120M/checkpoint-59000",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 59835,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008356313194618534,
|
| 13 |
+
"grad_norm": 35.479248046875,
|
| 14 |
+
"learning_rate": 4.96e-05,
|
| 15 |
+
"loss": 4.5691,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016712626389237067,
|
| 20 |
+
"grad_norm": 21.039249420166016,
|
| 21 |
+
"learning_rate": 4.958203421252212e-05,
|
| 22 |
+
"loss": 4.1783,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.016712626389237067,
|
| 27 |
+
"eval_loss": 4.322763442993164,
|
| 28 |
+
"eval_runtime": 22.3501,
|
| 29 |
+
"eval_samples_per_second": 199.641,
|
| 30 |
+
"eval_steps_per_second": 24.966,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.025068939583855605,
|
| 35 |
+
"grad_norm": 18.14427947998047,
|
| 36 |
+
"learning_rate": 4.916069773320974e-05,
|
| 37 |
+
"loss": 4.0526,
|
| 38 |
+
"step": 1500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.033425252778474135,
|
| 42 |
+
"grad_norm": 16.241540908813477,
|
| 43 |
+
"learning_rate": 4.874020392685599e-05,
|
| 44 |
+
"loss": 3.9836,
|
| 45 |
+
"step": 2000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.033425252778474135,
|
| 49 |
+
"eval_loss": 4.129116058349609,
|
| 50 |
+
"eval_runtime": 22.3714,
|
| 51 |
+
"eval_samples_per_second": 199.451,
|
| 52 |
+
"eval_steps_per_second": 24.943,
|
| 53 |
+
"step": 2000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.04178156597309267,
|
| 57 |
+
"grad_norm": 18.163301467895508,
|
| 58 |
+
"learning_rate": 4.8318867447543606e-05,
|
| 59 |
+
"loss": 3.8827,
|
| 60 |
+
"step": 2500
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.05013787916771121,
|
| 64 |
+
"grad_norm": 24.604848861694336,
|
| 65 |
+
"learning_rate": 4.789753096823123e-05,
|
| 66 |
+
"loss": 3.7892,
|
| 67 |
+
"step": 3000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.05013787916771121,
|
| 71 |
+
"eval_loss": 4.018996238708496,
|
| 72 |
+
"eval_runtime": 22.3789,
|
| 73 |
+
"eval_samples_per_second": 199.384,
|
| 74 |
+
"eval_steps_per_second": 24.934,
|
| 75 |
+
"step": 3000
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.05849419236232974,
|
| 79 |
+
"grad_norm": 11.950243949890137,
|
| 80 |
+
"learning_rate": 4.747619448891885e-05,
|
| 81 |
+
"loss": 3.8354,
|
| 82 |
+
"step": 3500
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.06685050555694827,
|
| 86 |
+
"grad_norm": 15.779219627380371,
|
| 87 |
+
"learning_rate": 4.705485800960647e-05,
|
| 88 |
+
"loss": 3.7895,
|
| 89 |
+
"step": 4000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.06685050555694827,
|
| 93 |
+
"eval_loss": 3.9454355239868164,
|
| 94 |
+
"eval_runtime": 22.2538,
|
| 95 |
+
"eval_samples_per_second": 200.505,
|
| 96 |
+
"eval_steps_per_second": 25.074,
|
| 97 |
+
"step": 4000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.0752068187515668,
|
| 101 |
+
"grad_norm": 18.874004364013672,
|
| 102 |
+
"learning_rate": 4.6633521530294095e-05,
|
| 103 |
+
"loss": 3.7398,
|
| 104 |
+
"step": 4500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.08356313194618534,
|
| 108 |
+
"grad_norm": 15.567193984985352,
|
| 109 |
+
"learning_rate": 4.621218505098172e-05,
|
| 110 |
+
"loss": 3.6895,
|
| 111 |
+
"step": 5000
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.08356313194618534,
|
| 115 |
+
"eval_loss": 3.897683620452881,
|
| 116 |
+
"eval_runtime": 22.3197,
|
| 117 |
+
"eval_samples_per_second": 199.913,
|
| 118 |
+
"eval_steps_per_second": 25.0,
|
| 119 |
+
"step": 5000
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.09191944514080387,
|
| 123 |
+
"grad_norm": 16.107986450195312,
|
| 124 |
+
"learning_rate": 4.579084857166934e-05,
|
| 125 |
+
"loss": 3.6716,
|
| 126 |
+
"step": 5500
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.10027575833542242,
|
| 130 |
+
"grad_norm": 13.824947357177734,
|
| 131 |
+
"learning_rate": 4.5369512092356955e-05,
|
| 132 |
+
"loss": 3.6242,
|
| 133 |
+
"step": 6000
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.10027575833542242,
|
| 137 |
+
"eval_loss": 3.839233636856079,
|
| 138 |
+
"eval_runtime": 22.2807,
|
| 139 |
+
"eval_samples_per_second": 200.263,
|
| 140 |
+
"eval_steps_per_second": 25.044,
|
| 141 |
+
"step": 6000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.10863207153004095,
|
| 145 |
+
"grad_norm": 16.347686767578125,
|
| 146 |
+
"learning_rate": 4.494817561304458e-05,
|
| 147 |
+
"loss": 3.5884,
|
| 148 |
+
"step": 6500
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.11698838472465949,
|
| 152 |
+
"grad_norm": 12.54311466217041,
|
| 153 |
+
"learning_rate": 4.4527681806690827e-05,
|
| 154 |
+
"loss": 3.6266,
|
| 155 |
+
"step": 7000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.11698838472465949,
|
| 159 |
+
"eval_loss": 3.802494525909424,
|
| 160 |
+
"eval_runtime": 22.3755,
|
| 161 |
+
"eval_samples_per_second": 199.415,
|
| 162 |
+
"eval_steps_per_second": 24.938,
|
| 163 |
+
"step": 7000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.12534469791927802,
|
| 167 |
+
"grad_norm": 14.918440818786621,
|
| 168 |
+
"learning_rate": 4.410634532737845e-05,
|
| 169 |
+
"loss": 3.6035,
|
| 170 |
+
"step": 7500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.13370101111389654,
|
| 174 |
+
"grad_norm": 14.448460578918457,
|
| 175 |
+
"learning_rate": 4.3685008848066064e-05,
|
| 176 |
+
"loss": 3.5471,
|
| 177 |
+
"step": 8000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.13370101111389654,
|
| 181 |
+
"eval_loss": 3.7752952575683594,
|
| 182 |
+
"eval_runtime": 22.2456,
|
| 183 |
+
"eval_samples_per_second": 200.579,
|
| 184 |
+
"eval_steps_per_second": 25.084,
|
| 185 |
+
"step": 8000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1420573243085151,
|
| 189 |
+
"grad_norm": 16.727405548095703,
|
| 190 |
+
"learning_rate": 4.326367236875369e-05,
|
| 191 |
+
"loss": 3.5496,
|
| 192 |
+
"step": 8500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1504136375031336,
|
| 196 |
+
"grad_norm": 12.798069953918457,
|
| 197 |
+
"learning_rate": 4.284233588944131e-05,
|
| 198 |
+
"loss": 3.5484,
|
| 199 |
+
"step": 9000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1504136375031336,
|
| 203 |
+
"eval_loss": 3.7489852905273438,
|
| 204 |
+
"eval_runtime": 22.3471,
|
| 205 |
+
"eval_samples_per_second": 199.668,
|
| 206 |
+
"eval_steps_per_second": 24.97,
|
| 207 |
+
"step": 9000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.15876995069775215,
|
| 211 |
+
"grad_norm": 11.953953742980957,
|
| 212 |
+
"learning_rate": 4.242099941012893e-05,
|
| 213 |
+
"loss": 3.5066,
|
| 214 |
+
"step": 9500
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.16712626389237067,
|
| 218 |
+
"grad_norm": 13.229384422302246,
|
| 219 |
+
"learning_rate": 4.1999662930816554e-05,
|
| 220 |
+
"loss": 3.496,
|
| 221 |
+
"step": 10000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.16712626389237067,
|
| 225 |
+
"eval_loss": 3.716583251953125,
|
| 226 |
+
"eval_runtime": 22.3577,
|
| 227 |
+
"eval_samples_per_second": 199.573,
|
| 228 |
+
"eval_steps_per_second": 24.958,
|
| 229 |
+
"step": 10000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.17548257708698922,
|
| 233 |
+
"grad_norm": 11.95522689819336,
|
| 234 |
+
"learning_rate": 4.1578326451504176e-05,
|
| 235 |
+
"loss": 3.462,
|
| 236 |
+
"step": 10500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.18383889028160774,
|
| 240 |
+
"grad_norm": 23.668384552001953,
|
| 241 |
+
"learning_rate": 4.11569899721918e-05,
|
| 242 |
+
"loss": 3.4485,
|
| 243 |
+
"step": 11000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.18383889028160774,
|
| 247 |
+
"eval_loss": 3.6793529987335205,
|
| 248 |
+
"eval_runtime": 22.3814,
|
| 249 |
+
"eval_samples_per_second": 199.362,
|
| 250 |
+
"eval_steps_per_second": 24.931,
|
| 251 |
+
"step": 11000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.1921952034762263,
|
| 255 |
+
"grad_norm": 17.76200294494629,
|
| 256 |
+
"learning_rate": 4.0735653492879414e-05,
|
| 257 |
+
"loss": 3.4631,
|
| 258 |
+
"step": 11500
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.20055151667084484,
|
| 262 |
+
"grad_norm": 10.677286148071289,
|
| 263 |
+
"learning_rate": 4.0314317013567036e-05,
|
| 264 |
+
"loss": 3.4498,
|
| 265 |
+
"step": 12000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.20055151667084484,
|
| 269 |
+
"eval_loss": 3.659281015396118,
|
| 270 |
+
"eval_runtime": 22.257,
|
| 271 |
+
"eval_samples_per_second": 200.476,
|
| 272 |
+
"eval_steps_per_second": 25.071,
|
| 273 |
+
"step": 12000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.20890782986546336,
|
| 277 |
+
"grad_norm": 15.71968936920166,
|
| 278 |
+
"learning_rate": 3.989298053425466e-05,
|
| 279 |
+
"loss": 3.4596,
|
| 280 |
+
"step": 12500
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.2172641430600819,
|
| 284 |
+
"grad_norm": 10.1958589553833,
|
| 285 |
+
"learning_rate": 3.947164405494228e-05,
|
| 286 |
+
"loss": 3.3978,
|
| 287 |
+
"step": 13000
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.2172641430600819,
|
| 291 |
+
"eval_loss": 3.649770736694336,
|
| 292 |
+
"eval_runtime": 22.2606,
|
| 293 |
+
"eval_samples_per_second": 200.444,
|
| 294 |
+
"eval_steps_per_second": 25.067,
|
| 295 |
+
"step": 13000
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.22562045625470042,
|
| 299 |
+
"grad_norm": 10.865114212036133,
|
| 300 |
+
"learning_rate": 3.90503075756299e-05,
|
| 301 |
+
"loss": 3.3866,
|
| 302 |
+
"step": 13500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.23397676944931897,
|
| 306 |
+
"grad_norm": 13.456110954284668,
|
| 307 |
+
"learning_rate": 3.8628971096317526e-05,
|
| 308 |
+
"loss": 3.3983,
|
| 309 |
+
"step": 14000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.23397676944931897,
|
| 313 |
+
"eval_loss": 3.6266236305236816,
|
| 314 |
+
"eval_runtime": 22.2738,
|
| 315 |
+
"eval_samples_per_second": 200.325,
|
| 316 |
+
"eval_steps_per_second": 25.052,
|
| 317 |
+
"step": 14000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.2423330826439375,
|
| 321 |
+
"grad_norm": 10.894895553588867,
|
| 322 |
+
"learning_rate": 3.820847728996377e-05,
|
| 323 |
+
"loss": 3.396,
|
| 324 |
+
"step": 14500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.25068939583855604,
|
| 328 |
+
"grad_norm": 15.169280052185059,
|
| 329 |
+
"learning_rate": 3.778714081065139e-05,
|
| 330 |
+
"loss": 3.3957,
|
| 331 |
+
"step": 15000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.25068939583855604,
|
| 335 |
+
"eval_loss": 3.598552703857422,
|
| 336 |
+
"eval_runtime": 22.3613,
|
| 337 |
+
"eval_samples_per_second": 199.541,
|
| 338 |
+
"eval_steps_per_second": 24.954,
|
| 339 |
+
"step": 15000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2590457090331746,
|
| 343 |
+
"grad_norm": 11.281633377075195,
|
| 344 |
+
"learning_rate": 3.736580433133901e-05,
|
| 345 |
+
"loss": 3.3797,
|
| 346 |
+
"step": 15500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2674020222277931,
|
| 350 |
+
"grad_norm": 14.883914947509766,
|
| 351 |
+
"learning_rate": 3.6944467852026635e-05,
|
| 352 |
+
"loss": 3.3148,
|
| 353 |
+
"step": 16000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2674020222277931,
|
| 357 |
+
"eval_loss": 3.595505475997925,
|
| 358 |
+
"eval_runtime": 22.34,
|
| 359 |
+
"eval_samples_per_second": 199.731,
|
| 360 |
+
"eval_steps_per_second": 24.978,
|
| 361 |
+
"step": 16000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.2757583354224116,
|
| 365 |
+
"grad_norm": 18.707447052001953,
|
| 366 |
+
"learning_rate": 3.652313137271425e-05,
|
| 367 |
+
"loss": 3.3563,
|
| 368 |
+
"step": 16500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2841146486170302,
|
| 372 |
+
"grad_norm": 17.429594039916992,
|
| 373 |
+
"learning_rate": 3.610179489340187e-05,
|
| 374 |
+
"loss": 3.3366,
|
| 375 |
+
"step": 17000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.2841146486170302,
|
| 379 |
+
"eval_loss": 3.5599985122680664,
|
| 380 |
+
"eval_runtime": 22.4605,
|
| 381 |
+
"eval_samples_per_second": 198.66,
|
| 382 |
+
"eval_steps_per_second": 24.844,
|
| 383 |
+
"step": 17000
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.2924709618116487,
|
| 387 |
+
"grad_norm": 13.94611644744873,
|
| 388 |
+
"learning_rate": 3.568130108704812e-05,
|
| 389 |
+
"loss": 3.287,
|
| 390 |
+
"step": 17500
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.3008272750062672,
|
| 394 |
+
"grad_norm": 14.198286056518555,
|
| 395 |
+
"learning_rate": 3.5260807280694364e-05,
|
| 396 |
+
"loss": 3.2916,
|
| 397 |
+
"step": 18000
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.3008272750062672,
|
| 401 |
+
"eval_loss": 3.549407720565796,
|
| 402 |
+
"eval_runtime": 22.3088,
|
| 403 |
+
"eval_samples_per_second": 200.011,
|
| 404 |
+
"eval_steps_per_second": 25.013,
|
| 405 |
+
"step": 18000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.30918358820088576,
|
| 409 |
+
"grad_norm": 14.446388244628906,
|
| 410 |
+
"learning_rate": 3.483947080138199e-05,
|
| 411 |
+
"loss": 3.2898,
|
| 412 |
+
"step": 18500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.3175399013955043,
|
| 416 |
+
"grad_norm": 11.602550506591797,
|
| 417 |
+
"learning_rate": 3.441813432206961e-05,
|
| 418 |
+
"loss": 3.2933,
|
| 419 |
+
"step": 19000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.3175399013955043,
|
| 423 |
+
"eval_loss": 3.531162977218628,
|
| 424 |
+
"eval_runtime": 22.2383,
|
| 425 |
+
"eval_samples_per_second": 200.645,
|
| 426 |
+
"eval_steps_per_second": 25.092,
|
| 427 |
+
"step": 19000
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.32589621459012286,
|
| 431 |
+
"grad_norm": 12.805480003356934,
|
| 432 |
+
"learning_rate": 3.3996797842757225e-05,
|
| 433 |
+
"loss": 3.3027,
|
| 434 |
+
"step": 19500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.33425252778474135,
|
| 438 |
+
"grad_norm": 15.020812034606934,
|
| 439 |
+
"learning_rate": 3.357546136344485e-05,
|
| 440 |
+
"loss": 3.2619,
|
| 441 |
+
"step": 20000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.33425252778474135,
|
| 445 |
+
"eval_loss": 3.5211801528930664,
|
| 446 |
+
"eval_runtime": 22.2737,
|
| 447 |
+
"eval_samples_per_second": 200.326,
|
| 448 |
+
"eval_steps_per_second": 25.052,
|
| 449 |
+
"step": 20000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.3426088409793599,
|
| 453 |
+
"grad_norm": 11.382076263427734,
|
| 454 |
+
"learning_rate": 3.3154967557091096e-05,
|
| 455 |
+
"loss": 3.293,
|
| 456 |
+
"step": 20500
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.35096515417397844,
|
| 460 |
+
"grad_norm": 8.293545722961426,
|
| 461 |
+
"learning_rate": 3.273363107777871e-05,
|
| 462 |
+
"loss": 3.2406,
|
| 463 |
+
"step": 21000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.35096515417397844,
|
| 467 |
+
"eval_loss": 3.5136260986328125,
|
| 468 |
+
"eval_runtime": 22.2375,
|
| 469 |
+
"eval_samples_per_second": 200.652,
|
| 470 |
+
"eval_steps_per_second": 25.093,
|
| 471 |
+
"step": 21000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.359321467368597,
|
| 475 |
+
"grad_norm": 9.583738327026367,
|
| 476 |
+
"learning_rate": 3.2312294598466334e-05,
|
| 477 |
+
"loss": 3.2416,
|
| 478 |
+
"step": 21500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.3676777805632155,
|
| 482 |
+
"grad_norm": 10.723130226135254,
|
| 483 |
+
"learning_rate": 3.1890958119153956e-05,
|
| 484 |
+
"loss": 3.2568,
|
| 485 |
+
"step": 22000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.3676777805632155,
|
| 489 |
+
"eval_loss": 3.491396427154541,
|
| 490 |
+
"eval_runtime": 22.2389,
|
| 491 |
+
"eval_samples_per_second": 200.639,
|
| 492 |
+
"eval_steps_per_second": 25.091,
|
| 493 |
+
"step": 22000
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.37603409375783403,
|
| 497 |
+
"grad_norm": 12.601049423217773,
|
| 498 |
+
"learning_rate": 3.14704643128002e-05,
|
| 499 |
+
"loss": 3.316,
|
| 500 |
+
"step": 22500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3843904069524526,
|
| 504 |
+
"grad_norm": 12.495054244995117,
|
| 505 |
+
"learning_rate": 3.104912783348782e-05,
|
| 506 |
+
"loss": 3.2242,
|
| 507 |
+
"step": 23000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.3843904069524526,
|
| 511 |
+
"eval_loss": 3.4799916744232178,
|
| 512 |
+
"eval_runtime": 22.2275,
|
| 513 |
+
"eval_samples_per_second": 200.742,
|
| 514 |
+
"eval_steps_per_second": 25.104,
|
| 515 |
+
"step": 23000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.3927467201470711,
|
| 519 |
+
"grad_norm": 10.878073692321777,
|
| 520 |
+
"learning_rate": 3.062779135417544e-05,
|
| 521 |
+
"loss": 3.2398,
|
| 522 |
+
"step": 23500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.4011030333416897,
|
| 526 |
+
"grad_norm": 11.56193733215332,
|
| 527 |
+
"learning_rate": 3.0206454874863066e-05,
|
| 528 |
+
"loss": 3.209,
|
| 529 |
+
"step": 24000
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.4011030333416897,
|
| 533 |
+
"eval_loss": 3.461331367492676,
|
| 534 |
+
"eval_runtime": 22.2718,
|
| 535 |
+
"eval_samples_per_second": 200.343,
|
| 536 |
+
"eval_steps_per_second": 25.054,
|
| 537 |
+
"step": 24000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.40945934653630817,
|
| 541 |
+
"grad_norm": 14.65617561340332,
|
| 542 |
+
"learning_rate": 2.9785118395550688e-05,
|
| 543 |
+
"loss": 3.2087,
|
| 544 |
+
"step": 24500
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.4178156597309267,
|
| 548 |
+
"grad_norm": 10.846756935119629,
|
| 549 |
+
"learning_rate": 2.936378191623831e-05,
|
| 550 |
+
"loss": 3.1671,
|
| 551 |
+
"step": 25000
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.4178156597309267,
|
| 555 |
+
"eval_loss": 3.447679042816162,
|
| 556 |
+
"eval_runtime": 22.2558,
|
| 557 |
+
"eval_samples_per_second": 200.487,
|
| 558 |
+
"eval_steps_per_second": 25.072,
|
| 559 |
+
"step": 25000
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.42617197292554526,
|
| 563 |
+
"grad_norm": 11.97311019897461,
|
| 564 |
+
"learning_rate": 2.894244543692593e-05,
|
| 565 |
+
"loss": 3.1925,
|
| 566 |
+
"step": 25500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.4345282861201638,
|
| 570 |
+
"grad_norm": 12.073387145996094,
|
| 571 |
+
"learning_rate": 2.852110895761355e-05,
|
| 572 |
+
"loss": 3.2222,
|
| 573 |
+
"step": 26000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.4345282861201638,
|
| 577 |
+
"eval_loss": 3.4386138916015625,
|
| 578 |
+
"eval_runtime": 22.247,
|
| 579 |
+
"eval_samples_per_second": 200.567,
|
| 580 |
+
"eval_steps_per_second": 25.082,
|
| 581 |
+
"step": 26000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.4428845993147823,
|
| 585 |
+
"grad_norm": 10.205495834350586,
|
| 586 |
+
"learning_rate": 2.8099772478301174e-05,
|
| 587 |
+
"loss": 3.1603,
|
| 588 |
+
"step": 26500
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.45124091250940085,
|
| 592 |
+
"grad_norm": 11.937155723571777,
|
| 593 |
+
"learning_rate": 2.7678435998988793e-05,
|
| 594 |
+
"loss": 3.1898,
|
| 595 |
+
"step": 27000
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.45124091250940085,
|
| 599 |
+
"eval_loss": 3.4313693046569824,
|
| 600 |
+
"eval_runtime": 22.2782,
|
| 601 |
+
"eval_samples_per_second": 200.286,
|
| 602 |
+
"eval_steps_per_second": 25.047,
|
| 603 |
+
"step": 27000
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.4595972257040194,
|
| 607 |
+
"grad_norm": 9.137228965759277,
|
| 608 |
+
"learning_rate": 2.725794219263504e-05,
|
| 609 |
+
"loss": 3.1701,
|
| 610 |
+
"step": 27500
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.46795353889863794,
|
| 614 |
+
"grad_norm": 16.823511123657227,
|
| 615 |
+
"learning_rate": 2.683660571332266e-05,
|
| 616 |
+
"loss": 3.142,
|
| 617 |
+
"step": 28000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.46795353889863794,
|
| 621 |
+
"eval_loss": 3.4176623821258545,
|
| 622 |
+
"eval_runtime": 22.2488,
|
| 623 |
+
"eval_samples_per_second": 200.55,
|
| 624 |
+
"eval_steps_per_second": 25.08,
|
| 625 |
+
"step": 28000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.47630985209325644,
|
| 629 |
+
"grad_norm": 8.703514099121094,
|
| 630 |
+
"learning_rate": 2.641526923401028e-05,
|
| 631 |
+
"loss": 3.167,
|
| 632 |
+
"step": 28500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.484666165287875,
|
| 636 |
+
"grad_norm": 10.721179962158203,
|
| 637 |
+
"learning_rate": 2.5993932754697902e-05,
|
| 638 |
+
"loss": 3.1781,
|
| 639 |
+
"step": 29000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.484666165287875,
|
| 643 |
+
"eval_loss": 3.410005569458008,
|
| 644 |
+
"eval_runtime": 22.2563,
|
| 645 |
+
"eval_samples_per_second": 200.482,
|
| 646 |
+
"eval_steps_per_second": 25.072,
|
| 647 |
+
"step": 29000
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.49302247848249353,
|
| 651 |
+
"grad_norm": 12.79716968536377,
|
| 652 |
+
"learning_rate": 2.5572596275385524e-05,
|
| 653 |
+
"loss": 3.1427,
|
| 654 |
+
"step": 29500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5013787916771121,
|
| 658 |
+
"grad_norm": 10.93630599975586,
|
| 659 |
+
"learning_rate": 2.5151259796073147e-05,
|
| 660 |
+
"loss": 3.1848,
|
| 661 |
+
"step": 30000
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5013787916771121,
|
| 665 |
+
"eval_loss": 3.393404483795166,
|
| 666 |
+
"eval_runtime": 22.2492,
|
| 667 |
+
"eval_samples_per_second": 200.547,
|
| 668 |
+
"eval_steps_per_second": 25.08,
|
| 669 |
+
"step": 30000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.5097351048717306,
|
| 673 |
+
"grad_norm": 13.065569877624512,
|
| 674 |
+
"learning_rate": 2.4729923316760766e-05,
|
| 675 |
+
"loss": 3.1423,
|
| 676 |
+
"step": 30500
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.5180914180663492,
|
| 680 |
+
"grad_norm": 13.116445541381836,
|
| 681 |
+
"learning_rate": 2.4308586837448388e-05,
|
| 682 |
+
"loss": 3.124,
|
| 683 |
+
"step": 31000
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.5180914180663492,
|
| 687 |
+
"eval_loss": 3.3824470043182373,
|
| 688 |
+
"eval_runtime": 22.2069,
|
| 689 |
+
"eval_samples_per_second": 200.929,
|
| 690 |
+
"eval_steps_per_second": 25.127,
|
| 691 |
+
"step": 31000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.5264477312609677,
|
| 695 |
+
"grad_norm": 13.007023811340332,
|
| 696 |
+
"learning_rate": 2.3888093031094634e-05,
|
| 697 |
+
"loss": 3.1545,
|
| 698 |
+
"step": 31500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.5348040444555862,
|
| 702 |
+
"grad_norm": 11.246246337890625,
|
| 703 |
+
"learning_rate": 2.3466756551782253e-05,
|
| 704 |
+
"loss": 3.1299,
|
| 705 |
+
"step": 32000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.5348040444555862,
|
| 709 |
+
"eval_loss": 3.3686602115631104,
|
| 710 |
+
"eval_runtime": 22.2456,
|
| 711 |
+
"eval_samples_per_second": 200.579,
|
| 712 |
+
"eval_steps_per_second": 25.084,
|
| 713 |
+
"step": 32000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.5431603576502048,
|
| 717 |
+
"grad_norm": 10.350507736206055,
|
| 718 |
+
"learning_rate": 2.3045420072469875e-05,
|
| 719 |
+
"loss": 3.1309,
|
| 720 |
+
"step": 32500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.5515166708448233,
|
| 724 |
+
"grad_norm": 11.70686149597168,
|
| 725 |
+
"learning_rate": 2.2624083593157497e-05,
|
| 726 |
+
"loss": 3.1039,
|
| 727 |
+
"step": 33000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.5515166708448233,
|
| 731 |
+
"eval_loss": 3.361196756362915,
|
| 732 |
+
"eval_runtime": 22.2263,
|
| 733 |
+
"eval_samples_per_second": 200.753,
|
| 734 |
+
"eval_steps_per_second": 25.105,
|
| 735 |
+
"step": 33000
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.5598729840394417,
|
| 739 |
+
"grad_norm": 14.55528736114502,
|
| 740 |
+
"learning_rate": 2.220358978680374e-05,
|
| 741 |
+
"loss": 3.1191,
|
| 742 |
+
"step": 33500
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.5682292972340603,
|
| 746 |
+
"grad_norm": 10.029934883117676,
|
| 747 |
+
"learning_rate": 2.178309598044999e-05,
|
| 748 |
+
"loss": 3.1253,
|
| 749 |
+
"step": 34000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.5682292972340603,
|
| 753 |
+
"eval_loss": 3.3478872776031494,
|
| 754 |
+
"eval_runtime": 25.1339,
|
| 755 |
+
"eval_samples_per_second": 177.529,
|
| 756 |
+
"eval_steps_per_second": 22.201,
|
| 757 |
+
"step": 34000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.5765856104286788,
|
| 761 |
+
"grad_norm": 12.41011905670166,
|
| 762 |
+
"learning_rate": 2.136175950113761e-05,
|
| 763 |
+
"loss": 3.1118,
|
| 764 |
+
"step": 34500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.5849419236232974,
|
| 768 |
+
"grad_norm": 7.660207748413086,
|
| 769 |
+
"learning_rate": 2.094042302182523e-05,
|
| 770 |
+
"loss": 3.0562,
|
| 771 |
+
"step": 35000
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.5849419236232974,
|
| 775 |
+
"eval_loss": 3.343930244445801,
|
| 776 |
+
"eval_runtime": 22.2454,
|
| 777 |
+
"eval_samples_per_second": 200.581,
|
| 778 |
+
"eval_steps_per_second": 25.084,
|
| 779 |
+
"step": 35000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5932982368179159,
|
| 783 |
+
"grad_norm": 14.761736869812012,
|
| 784 |
+
"learning_rate": 2.0519086542512852e-05,
|
| 785 |
+
"loss": 3.1209,
|
| 786 |
+
"step": 35500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.6016545500125344,
|
| 790 |
+
"grad_norm": 13.07590103149414,
|
| 791 |
+
"learning_rate": 2.0097750063200475e-05,
|
| 792 |
+
"loss": 3.0823,
|
| 793 |
+
"step": 36000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6016545500125344,
|
| 797 |
+
"eval_loss": 3.3252499103546143,
|
| 798 |
+
"eval_runtime": 22.2801,
|
| 799 |
+
"eval_samples_per_second": 200.268,
|
| 800 |
+
"eval_steps_per_second": 25.045,
|
| 801 |
+
"step": 36000
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.610010863207153,
|
| 805 |
+
"grad_norm": 11.221796989440918,
|
| 806 |
+
"learning_rate": 1.9677256256846717e-05,
|
| 807 |
+
"loss": 3.0989,
|
| 808 |
+
"step": 36500
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6183671764017715,
|
| 812 |
+
"grad_norm": 12.109387397766113,
|
| 813 |
+
"learning_rate": 1.925591977753434e-05,
|
| 814 |
+
"loss": 3.0864,
|
| 815 |
+
"step": 37000
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.6183671764017715,
|
| 819 |
+
"eval_loss": 3.313296318054199,
|
| 820 |
+
"eval_runtime": 22.2819,
|
| 821 |
+
"eval_samples_per_second": 200.252,
|
| 822 |
+
"eval_steps_per_second": 25.043,
|
| 823 |
+
"step": 37000
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.6267234895963901,
|
| 827 |
+
"grad_norm": 10.61958122253418,
|
| 828 |
+
"learning_rate": 1.8834583298221962e-05,
|
| 829 |
+
"loss": 3.0931,
|
| 830 |
+
"step": 37500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.6350798027910086,
|
| 834 |
+
"grad_norm": 12.216238021850586,
|
| 835 |
+
"learning_rate": 1.8413246818909584e-05,
|
| 836 |
+
"loss": 3.1024,
|
| 837 |
+
"step": 38000
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.6350798027910086,
|
| 841 |
+
"eval_loss": 3.300213098526001,
|
| 842 |
+
"eval_runtime": 22.3151,
|
| 843 |
+
"eval_samples_per_second": 199.955,
|
| 844 |
+
"eval_steps_per_second": 25.006,
|
| 845 |
+
"step": 38000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.6434361159856271,
|
| 849 |
+
"grad_norm": 11.379255294799805,
|
| 850 |
+
"learning_rate": 1.7991910339597203e-05,
|
| 851 |
+
"loss": 3.021,
|
| 852 |
+
"step": 38500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.6517924291802457,
|
| 856 |
+
"grad_norm": 12.111197471618652,
|
| 857 |
+
"learning_rate": 1.7570573860284825e-05,
|
| 858 |
+
"loss": 3.0478,
|
| 859 |
+
"step": 39000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.6517924291802457,
|
| 863 |
+
"eval_loss": 3.294560432434082,
|
| 864 |
+
"eval_runtime": 22.2566,
|
| 865 |
+
"eval_samples_per_second": 200.479,
|
| 866 |
+
"eval_steps_per_second": 25.071,
|
| 867 |
+
"step": 39000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.6601487423748642,
|
| 871 |
+
"grad_norm": 14.204049110412598,
|
| 872 |
+
"learning_rate": 1.7149237380972448e-05,
|
| 873 |
+
"loss": 3.0032,
|
| 874 |
+
"step": 39500
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.6685050555694827,
|
| 878 |
+
"grad_norm": 11.631662368774414,
|
| 879 |
+
"learning_rate": 1.6727900901660067e-05,
|
| 880 |
+
"loss": 3.0655,
|
| 881 |
+
"step": 40000
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.6685050555694827,
|
| 885 |
+
"eval_loss": 3.285043239593506,
|
| 886 |
+
"eval_runtime": 22.2429,
|
| 887 |
+
"eval_samples_per_second": 200.603,
|
| 888 |
+
"eval_steps_per_second": 25.087,
|
| 889 |
+
"step": 40000
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.6768613687641013,
|
| 893 |
+
"grad_norm": 13.394169807434082,
|
| 894 |
+
"learning_rate": 1.6307407095306312e-05,
|
| 895 |
+
"loss": 3.0097,
|
| 896 |
+
"step": 40500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.6852176819587198,
|
| 900 |
+
"grad_norm": 14.214789390563965,
|
| 901 |
+
"learning_rate": 1.5886070615993935e-05,
|
| 902 |
+
"loss": 3.0313,
|
| 903 |
+
"step": 41000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.6852176819587198,
|
| 907 |
+
"eval_loss": 3.277559280395508,
|
| 908 |
+
"eval_runtime": 22.242,
|
| 909 |
+
"eval_samples_per_second": 200.611,
|
| 910 |
+
"eval_steps_per_second": 25.088,
|
| 911 |
+
"step": 41000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.6935739951533384,
|
| 915 |
+
"grad_norm": 8.208776473999023,
|
| 916 |
+
"learning_rate": 1.5464734136681554e-05,
|
| 917 |
+
"loss": 2.9826,
|
| 918 |
+
"step": 41500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.7019303083479569,
|
| 922 |
+
"grad_norm": 10.259654998779297,
|
| 923 |
+
"learning_rate": 1.50442403303278e-05,
|
| 924 |
+
"loss": 3.0034,
|
| 925 |
+
"step": 42000
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.7019303083479569,
|
| 929 |
+
"eval_loss": 3.2744297981262207,
|
| 930 |
+
"eval_runtime": 22.2597,
|
| 931 |
+
"eval_samples_per_second": 200.452,
|
| 932 |
+
"eval_steps_per_second": 25.068,
|
| 933 |
+
"step": 42000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.7102866215425754,
|
| 937 |
+
"grad_norm": 14.066951751708984,
|
| 938 |
+
"learning_rate": 1.462290385101542e-05,
|
| 939 |
+
"loss": 2.9843,
|
| 940 |
+
"step": 42500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.718642934737194,
|
| 944 |
+
"grad_norm": 16.31036376953125,
|
| 945 |
+
"learning_rate": 1.4201567371703042e-05,
|
| 946 |
+
"loss": 3.0232,
|
| 947 |
+
"step": 43000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.718642934737194,
|
| 951 |
+
"eval_loss": 3.2568132877349854,
|
| 952 |
+
"eval_runtime": 22.2183,
|
| 953 |
+
"eval_samples_per_second": 200.825,
|
| 954 |
+
"eval_steps_per_second": 25.114,
|
| 955 |
+
"step": 43000
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.7269992479318125,
|
| 959 |
+
"grad_norm": 12.592016220092773,
|
| 960 |
+
"learning_rate": 1.3780230892390663e-05,
|
| 961 |
+
"loss": 2.9779,
|
| 962 |
+
"step": 43500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.735355561126431,
|
| 966 |
+
"grad_norm": 11.640332221984863,
|
| 967 |
+
"learning_rate": 1.3358894413078285e-05,
|
| 968 |
+
"loss": 2.9875,
|
| 969 |
+
"step": 44000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.735355561126431,
|
| 973 |
+
"eval_loss": 3.2479593753814697,
|
| 974 |
+
"eval_runtime": 22.2574,
|
| 975 |
+
"eval_samples_per_second": 200.473,
|
| 976 |
+
"eval_steps_per_second": 25.07,
|
| 977 |
+
"step": 44000
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.7437118743210496,
|
| 981 |
+
"grad_norm": 18.809364318847656,
|
| 982 |
+
"learning_rate": 1.2937557933765906e-05,
|
| 983 |
+
"loss": 3.014,
|
| 984 |
+
"step": 44500
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.7520681875156681,
|
| 988 |
+
"grad_norm": 10.02562141418457,
|
| 989 |
+
"learning_rate": 1.2516221454453528e-05,
|
| 990 |
+
"loss": 2.9864,
|
| 991 |
+
"step": 45000
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.7520681875156681,
|
| 995 |
+
"eval_loss": 3.23964262008667,
|
| 996 |
+
"eval_runtime": 22.2529,
|
| 997 |
+
"eval_samples_per_second": 200.513,
|
| 998 |
+
"eval_steps_per_second": 25.075,
|
| 999 |
+
"step": 45000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.7604245007102867,
|
| 1003 |
+
"grad_norm": 14.984621047973633,
|
| 1004 |
+
"learning_rate": 1.2095727648099774e-05,
|
| 1005 |
+
"loss": 3.0169,
|
| 1006 |
+
"step": 45500
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.7687808139049052,
|
| 1010 |
+
"grad_norm": 14.823599815368652,
|
| 1011 |
+
"learning_rate": 1.1674391168787393e-05,
|
| 1012 |
+
"loss": 2.9889,
|
| 1013 |
+
"step": 46000
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.7687808139049052,
|
| 1017 |
+
"eval_loss": 3.2315359115600586,
|
| 1018 |
+
"eval_runtime": 22.2439,
|
| 1019 |
+
"eval_samples_per_second": 200.594,
|
| 1020 |
+
"eval_steps_per_second": 25.086,
|
| 1021 |
+
"step": 46000
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 0.7771371270995237,
|
| 1025 |
+
"grad_norm": 15.036163330078125,
|
| 1026 |
+
"learning_rate": 1.1253054689475015e-05,
|
| 1027 |
+
"loss": 2.9498,
|
| 1028 |
+
"step": 46500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.7854934402941423,
|
| 1032 |
+
"grad_norm": 11.967373847961426,
|
| 1033 |
+
"learning_rate": 1.0831718210162636e-05,
|
| 1034 |
+
"loss": 2.9473,
|
| 1035 |
+
"step": 47000
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.7854934402941423,
|
| 1039 |
+
"eval_loss": 3.2260513305664062,
|
| 1040 |
+
"eval_runtime": 22.2171,
|
| 1041 |
+
"eval_samples_per_second": 200.836,
|
| 1042 |
+
"eval_steps_per_second": 25.116,
|
| 1043 |
+
"step": 47000
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.7938497534887607,
|
| 1047 |
+
"grad_norm": 16.133859634399414,
|
| 1048 |
+
"learning_rate": 1.0410381730850258e-05,
|
| 1049 |
+
"loss": 2.9733,
|
| 1050 |
+
"step": 47500
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.8022060666833793,
|
| 1054 |
+
"grad_norm": 9.699966430664062,
|
| 1055 |
+
"learning_rate": 9.989045251537879e-06,
|
| 1056 |
+
"loss": 2.9981,
|
| 1057 |
+
"step": 48000
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.8022060666833793,
|
| 1061 |
+
"eval_loss": 3.219076156616211,
|
| 1062 |
+
"eval_runtime": 22.2412,
|
| 1063 |
+
"eval_samples_per_second": 200.618,
|
| 1064 |
+
"eval_steps_per_second": 25.089,
|
| 1065 |
+
"step": 48000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.8105623798779978,
|
| 1069 |
+
"grad_norm": 12.452173233032227,
|
| 1070 |
+
"learning_rate": 9.5677087722255e-06,
|
| 1071 |
+
"loss": 2.9572,
|
| 1072 |
+
"step": 48500
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.8189186930726163,
|
| 1076 |
+
"grad_norm": 10.909358024597168,
|
| 1077 |
+
"learning_rate": 9.14637229291312e-06,
|
| 1078 |
+
"loss": 2.9086,
|
| 1079 |
+
"step": 49000
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.8189186930726163,
|
| 1083 |
+
"eval_loss": 3.2129132747650146,
|
| 1084 |
+
"eval_runtime": 22.2484,
|
| 1085 |
+
"eval_samples_per_second": 200.554,
|
| 1086 |
+
"eval_steps_per_second": 25.08,
|
| 1087 |
+
"step": 49000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.8272750062672349,
|
| 1091 |
+
"grad_norm": 15.663894653320312,
|
| 1092 |
+
"learning_rate": 8.725878486559366e-06,
|
| 1093 |
+
"loss": 2.9755,
|
| 1094 |
+
"step": 49500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.8356313194618534,
|
| 1098 |
+
"grad_norm": 9.793340682983398,
|
| 1099 |
+
"learning_rate": 8.304542007246988e-06,
|
| 1100 |
+
"loss": 2.9533,
|
| 1101 |
+
"step": 50000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.8356313194618534,
|
| 1105 |
+
"eval_loss": 3.204308271408081,
|
| 1106 |
+
"eval_runtime": 22.2781,
|
| 1107 |
+
"eval_samples_per_second": 200.286,
|
| 1108 |
+
"eval_steps_per_second": 25.047,
|
| 1109 |
+
"step": 50000
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.8439876326564719,
|
| 1113 |
+
"grad_norm": 13.584871292114258,
|
| 1114 |
+
"learning_rate": 7.883205527934609e-06,
|
| 1115 |
+
"loss": 2.9323,
|
| 1116 |
+
"step": 50500
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.8523439458510905,
|
| 1120 |
+
"grad_norm": 17.01344871520996,
|
| 1121 |
+
"learning_rate": 7.4618690486222304e-06,
|
| 1122 |
+
"loss": 2.9298,
|
| 1123 |
+
"step": 51000
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.8523439458510905,
|
| 1127 |
+
"eval_loss": 3.1974070072174072,
|
| 1128 |
+
"eval_runtime": 22.2555,
|
| 1129 |
+
"eval_samples_per_second": 200.489,
|
| 1130 |
+
"eval_steps_per_second": 25.072,
|
| 1131 |
+
"step": 51000
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.860700259045709,
|
| 1135 |
+
"grad_norm": 11.150009155273438,
|
| 1136 |
+
"learning_rate": 7.041375242268476e-06,
|
| 1137 |
+
"loss": 2.9175,
|
| 1138 |
+
"step": 51500
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.8690565722403276,
|
| 1142 |
+
"grad_norm": 10.69240951538086,
|
| 1143 |
+
"learning_rate": 6.620038762956098e-06,
|
| 1144 |
+
"loss": 2.9156,
|
| 1145 |
+
"step": 52000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.8690565722403276,
|
| 1149 |
+
"eval_loss": 3.192389488220215,
|
| 1150 |
+
"eval_runtime": 22.2526,
|
| 1151 |
+
"eval_samples_per_second": 200.516,
|
| 1152 |
+
"eval_steps_per_second": 25.076,
|
| 1153 |
+
"step": 52000
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.8774128854349461,
|
| 1157 |
+
"grad_norm": 12.42056655883789,
|
| 1158 |
+
"learning_rate": 6.198702283643718e-06,
|
| 1159 |
+
"loss": 2.8909,
|
| 1160 |
+
"step": 52500
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.8857691986295646,
|
| 1164 |
+
"grad_norm": 11.83171558380127,
|
| 1165 |
+
"learning_rate": 5.77736580433134e-06,
|
| 1166 |
+
"loss": 2.9452,
|
| 1167 |
+
"step": 53000
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.8857691986295646,
|
| 1171 |
+
"eval_loss": 3.186549425125122,
|
| 1172 |
+
"eval_runtime": 22.2935,
|
| 1173 |
+
"eval_samples_per_second": 200.148,
|
| 1174 |
+
"eval_steps_per_second": 25.03,
|
| 1175 |
+
"step": 53000
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.8941255118241832,
|
| 1179 |
+
"grad_norm": 17.534683227539062,
|
| 1180 |
+
"learning_rate": 5.3568719979775856e-06,
|
| 1181 |
+
"loss": 2.9173,
|
| 1182 |
+
"step": 53500
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.9024818250188017,
|
| 1186 |
+
"grad_norm": 13.407784461975098,
|
| 1187 |
+
"learning_rate": 4.935535518665206e-06,
|
| 1188 |
+
"loss": 2.959,
|
| 1189 |
+
"step": 54000
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.9024818250188017,
|
| 1193 |
+
"eval_loss": 3.1784276962280273,
|
| 1194 |
+
"eval_runtime": 22.2696,
|
| 1195 |
+
"eval_samples_per_second": 200.363,
|
| 1196 |
+
"eval_steps_per_second": 25.057,
|
| 1197 |
+
"step": 54000
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.9108381382134202,
|
| 1201 |
+
"grad_norm": 14.451789855957031,
|
| 1202 |
+
"learning_rate": 4.514199039352828e-06,
|
| 1203 |
+
"loss": 2.9441,
|
| 1204 |
+
"step": 54500
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 0.9191944514080388,
|
| 1208 |
+
"grad_norm": 11.849891662597656,
|
| 1209 |
+
"learning_rate": 4.092862560040449e-06,
|
| 1210 |
+
"loss": 2.8955,
|
| 1211 |
+
"step": 55000
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.9191944514080388,
|
| 1215 |
+
"eval_loss": 3.1731808185577393,
|
| 1216 |
+
"eval_runtime": 22.2282,
|
| 1217 |
+
"eval_samples_per_second": 200.736,
|
| 1218 |
+
"eval_steps_per_second": 25.103,
|
| 1219 |
+
"step": 55000
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 0.9275507646026573,
|
| 1223 |
+
"grad_norm": 13.188152313232422,
|
| 1224 |
+
"learning_rate": 3.6715260807280694e-06,
|
| 1225 |
+
"loss": 2.889,
|
| 1226 |
+
"step": 55500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.9359070777972759,
|
| 1230 |
+
"grad_norm": 16.554080963134766,
|
| 1231 |
+
"learning_rate": 3.250189601415691e-06,
|
| 1232 |
+
"loss": 2.9453,
|
| 1233 |
+
"step": 56000
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.9359070777972759,
|
| 1237 |
+
"eval_loss": 3.167375087738037,
|
| 1238 |
+
"eval_runtime": 22.2342,
|
| 1239 |
+
"eval_samples_per_second": 200.682,
|
| 1240 |
+
"eval_steps_per_second": 25.096,
|
| 1241 |
+
"step": 56000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.9442633909918944,
|
| 1245 |
+
"grad_norm": 11.261225700378418,
|
| 1246 |
+
"learning_rate": 2.828853122103312e-06,
|
| 1247 |
+
"loss": 2.9546,
|
| 1248 |
+
"step": 56500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.9526197041865129,
|
| 1252 |
+
"grad_norm": 9.682096481323242,
|
| 1253 |
+
"learning_rate": 2.407516642790933e-06,
|
| 1254 |
+
"loss": 2.8689,
|
| 1255 |
+
"step": 57000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.9526197041865129,
|
| 1259 |
+
"eval_loss": 3.1633212566375732,
|
| 1260 |
+
"eval_runtime": 22.2444,
|
| 1261 |
+
"eval_samples_per_second": 200.59,
|
| 1262 |
+
"eval_steps_per_second": 25.085,
|
| 1263 |
+
"step": 57000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.9609760173811315,
|
| 1267 |
+
"grad_norm": 10.658622741699219,
|
| 1268 |
+
"learning_rate": 1.9870228364371787e-06,
|
| 1269 |
+
"loss": 2.8708,
|
| 1270 |
+
"step": 57500
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.96933233057575,
|
| 1274 |
+
"grad_norm": 11.864119529724121,
|
| 1275 |
+
"learning_rate": 1.5656863571248002e-06,
|
| 1276 |
+
"loss": 2.8988,
|
| 1277 |
+
"step": 58000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.96933233057575,
|
| 1281 |
+
"eval_loss": 3.1584391593933105,
|
| 1282 |
+
"eval_runtime": 22.2736,
|
| 1283 |
+
"eval_samples_per_second": 200.327,
|
| 1284 |
+
"eval_steps_per_second": 25.052,
|
| 1285 |
+
"step": 58000
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.9776886437703685,
|
| 1289 |
+
"grad_norm": 9.938982009887695,
|
| 1290 |
+
"learning_rate": 1.144349877812421e-06,
|
| 1291 |
+
"loss": 2.9186,
|
| 1292 |
+
"step": 58500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.9860449569649871,
|
| 1296 |
+
"grad_norm": 12.801502227783203,
|
| 1297 |
+
"learning_rate": 7.230133985000422e-07,
|
| 1298 |
+
"loss": 2.8888,
|
| 1299 |
+
"step": 59000
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.9860449569649871,
|
| 1303 |
+
"eval_loss": 3.157134532928467,
|
| 1304 |
+
"eval_runtime": 22.2821,
|
| 1305 |
+
"eval_samples_per_second": 200.251,
|
| 1306 |
+
"eval_steps_per_second": 25.043,
|
| 1307 |
+
"step": 59000
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.9944012701596056,
|
| 1311 |
+
"grad_norm": 16.9565372467041,
|
| 1312 |
+
"learning_rate": 3.025195921462881e-07,
|
| 1313 |
+
"loss": 2.8857,
|
| 1314 |
+
"step": 59500
|
| 1315 |
+
}
|
| 1316 |
+
],
|
| 1317 |
+
"logging_steps": 500,
|
| 1318 |
+
"max_steps": 59835,
|
| 1319 |
+
"num_input_tokens_seen": 0,
|
| 1320 |
+
"num_train_epochs": 1,
|
| 1321 |
+
"save_steps": 2000,
|
| 1322 |
+
"stateful_callbacks": {
|
| 1323 |
+
"TrainerControl": {
|
| 1324 |
+
"args": {
|
| 1325 |
+
"should_epoch_stop": false,
|
| 1326 |
+
"should_evaluate": false,
|
| 1327 |
+
"should_log": false,
|
| 1328 |
+
"should_save": true,
|
| 1329 |
+
"should_training_stop": true
|
| 1330 |
+
},
|
| 1331 |
+
"attributes": {}
|
| 1332 |
+
}
|
| 1333 |
+
},
|
| 1334 |
+
"total_flos": 2.943723113325527e+17,
|
| 1335 |
+
"train_batch_size": 8,
|
| 1336 |
+
"trial_name": null,
|
| 1337 |
+
"trial_params": null
|
| 1338 |
+
}
|
custom-tokenizer/full-finetuning/LLaMmlein_120M/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f88c9a573a75529b8b6e9fd0abb8f8716c5a97badf4da729f6e39e4f77539b11
|
| 3 |
+
size 5368
|
custom-tokenizer/full-finetuning/german-gpt2/config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "dbmdz/german-gpt2",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.0,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"embd_pdrop": 0.0,
|
| 10 |
+
"eos_token_id": 50256,
|
| 11 |
+
"gradient_checkpointing": false,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"layer_norm_epsilon": 1e-05,
|
| 14 |
+
"model_type": "gpt2",
|
| 15 |
+
"n_ctx": 1024,
|
| 16 |
+
"n_embd": 768,
|
| 17 |
+
"n_head": 12,
|
| 18 |
+
"n_inner": null,
|
| 19 |
+
"n_layer": 12,
|
| 20 |
+
"n_positions": 1024,
|
| 21 |
+
"reorder_and_upcast_attn": false,
|
| 22 |
+
"resid_pdrop": 0.0,
|
| 23 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 24 |
+
"scale_attn_weights": true,
|
| 25 |
+
"summary_activation": null,
|
| 26 |
+
"summary_first_dropout": 0.1,
|
| 27 |
+
"summary_proj_to_labels": true,
|
| 28 |
+
"summary_type": "cls_index",
|
| 29 |
+
"summary_use_proj": true,
|
| 30 |
+
"task_specific_params": {
|
| 31 |
+
"text-generation": {
|
| 32 |
+
"do_sample": true,
|
| 33 |
+
"max_length": 50
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"torch_dtype": "float32",
|
| 37 |
+
"transformers_version": "4.47.1",
|
| 38 |
+
"use_cache": true,
|
| 39 |
+
"vocab_size": 50266
|
| 40 |
+
}
|
custom-tokenizer/full-finetuning/german-gpt2/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.47.1"
|
| 6 |
+
}
|
custom-tokenizer/full-finetuning/german-gpt2/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6eb02a452b7bcd4d795b4683e55cc8658721a3944d1ab17594f817ff07d3c06d
|
| 3 |
+
size 497801856
|
custom-tokenizer/full-finetuning/german-gpt2/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3962fa444f5a31e9641be7ed55d6a2cc3604aa41ef2757f52f75bd66f0075a6
|
| 3 |
+
size 995697594
|
custom-tokenizer/full-finetuning/german-gpt2/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3bc1ac846859ad13fd356d3ed41fd20834205a7e8764e4198636863fca3f64c
|
| 3 |
+
size 14244
|
custom-tokenizer/full-finetuning/german-gpt2/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23f99f08d1d78974aae2a3a91c7471d7ba832f92d524504de2b7948fa62af5d8
|
| 3 |
+
size 1064
|
custom-tokenizer/full-finetuning/german-gpt2/trainer_state.json
ADDED
|
@@ -0,0 +1,1338 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.311403751373291,
|
| 3 |
+
"best_model_checkpoint": "./models/custom-tokenizer/full-finetuning/german-gpt2/checkpoint-59000",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 59835,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008356313194618534,
|
| 13 |
+
"grad_norm": 7.739098072052002,
|
| 14 |
+
"learning_rate": 4.97e-05,
|
| 15 |
+
"loss": 5.7212,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016712626389237067,
|
| 20 |
+
"grad_norm": 6.011812686920166,
|
| 21 |
+
"learning_rate": 4.95811915395635e-05,
|
| 22 |
+
"loss": 4.8525,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.016712626389237067,
|
| 27 |
+
"eval_loss": 4.5701584815979,
|
| 28 |
+
"eval_runtime": 34.6716,
|
| 29 |
+
"eval_samples_per_second": 128.693,
|
| 30 |
+
"eval_steps_per_second": 16.094,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.025068939583855605,
|
| 35 |
+
"grad_norm": 5.455048561096191,
|
| 36 |
+
"learning_rate": 4.915985506025112e-05,
|
| 37 |
+
"loss": 4.6135,
|
| 38 |
+
"step": 1500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.033425252778474135,
|
| 42 |
+
"grad_norm": 7.002511024475098,
|
| 43 |
+
"learning_rate": 4.873851858093874e-05,
|
| 44 |
+
"loss": 4.4939,
|
| 45 |
+
"step": 2000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.033425252778474135,
|
| 49 |
+
"eval_loss": 4.25942325592041,
|
| 50 |
+
"eval_runtime": 34.6626,
|
| 51 |
+
"eval_samples_per_second": 128.727,
|
| 52 |
+
"eval_steps_per_second": 16.098,
|
| 53 |
+
"step": 2000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.04178156597309267,
|
| 57 |
+
"grad_norm": 6.586827754974365,
|
| 58 |
+
"learning_rate": 4.8317182101626365e-05,
|
| 59 |
+
"loss": 4.3682,
|
| 60 |
+
"step": 2500
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.05013787916771121,
|
| 64 |
+
"grad_norm": 5.179999351501465,
|
| 65 |
+
"learning_rate": 4.789584562231398e-05,
|
| 66 |
+
"loss": 4.2088,
|
| 67 |
+
"step": 3000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.05013787916771121,
|
| 71 |
+
"eval_loss": 4.1048455238342285,
|
| 72 |
+
"eval_runtime": 34.6948,
|
| 73 |
+
"eval_samples_per_second": 128.607,
|
| 74 |
+
"eval_steps_per_second": 16.083,
|
| 75 |
+
"step": 3000
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.05849419236232974,
|
| 79 |
+
"grad_norm": 4.009952545166016,
|
| 80 |
+
"learning_rate": 4.74745091430016e-05,
|
| 81 |
+
"loss": 4.2534,
|
| 82 |
+
"step": 3500
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.06685050555694827,
|
| 86 |
+
"grad_norm": 5.067552089691162,
|
| 87 |
+
"learning_rate": 4.7053172663689226e-05,
|
| 88 |
+
"loss": 4.2047,
|
| 89 |
+
"step": 4000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.06685050555694827,
|
| 93 |
+
"eval_loss": 4.001619815826416,
|
| 94 |
+
"eval_runtime": 34.7045,
|
| 95 |
+
"eval_samples_per_second": 128.571,
|
| 96 |
+
"eval_steps_per_second": 16.079,
|
| 97 |
+
"step": 4000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.0752068187515668,
|
| 101 |
+
"grad_norm": 8.432042121887207,
|
| 102 |
+
"learning_rate": 4.663183618437685e-05,
|
| 103 |
+
"loss": 4.1514,
|
| 104 |
+
"step": 4500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.08356313194618534,
|
| 108 |
+
"grad_norm": 5.663743019104004,
|
| 109 |
+
"learning_rate": 4.621049970506447e-05,
|
| 110 |
+
"loss": 4.0911,
|
| 111 |
+
"step": 5000
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.08356313194618534,
|
| 115 |
+
"eval_loss": 3.927936553955078,
|
| 116 |
+
"eval_runtime": 34.7705,
|
| 117 |
+
"eval_samples_per_second": 128.327,
|
| 118 |
+
"eval_steps_per_second": 16.048,
|
| 119 |
+
"step": 5000
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.09191944514080387,
|
| 123 |
+
"grad_norm": 6.105470180511475,
|
| 124 |
+
"learning_rate": 4.5789163225752086e-05,
|
| 125 |
+
"loss": 4.0603,
|
| 126 |
+
"step": 5500
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.10027575833542242,
|
| 130 |
+
"grad_norm": 4.524149417877197,
|
| 131 |
+
"learning_rate": 4.536782674643971e-05,
|
| 132 |
+
"loss": 4.0241,
|
| 133 |
+
"step": 6000
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.10027575833542242,
|
| 137 |
+
"eval_loss": 3.868943691253662,
|
| 138 |
+
"eval_runtime": 34.7126,
|
| 139 |
+
"eval_samples_per_second": 128.541,
|
| 140 |
+
"eval_steps_per_second": 16.075,
|
| 141 |
+
"step": 6000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.10863207153004095,
|
| 145 |
+
"grad_norm": 5.74066686630249,
|
| 146 |
+
"learning_rate": 4.494649026712733e-05,
|
| 147 |
+
"loss": 3.9523,
|
| 148 |
+
"step": 6500
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.11698838472465949,
|
| 152 |
+
"grad_norm": 5.0315046310424805,
|
| 153 |
+
"learning_rate": 4.452599646077358e-05,
|
| 154 |
+
"loss": 3.9988,
|
| 155 |
+
"step": 7000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.11698838472465949,
|
| 159 |
+
"eval_loss": 3.817981719970703,
|
| 160 |
+
"eval_runtime": 34.7005,
|
| 161 |
+
"eval_samples_per_second": 128.586,
|
| 162 |
+
"eval_steps_per_second": 16.08,
|
| 163 |
+
"step": 7000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.12534469791927802,
|
| 167 |
+
"grad_norm": 6.331228733062744,
|
| 168 |
+
"learning_rate": 4.41046599814612e-05,
|
| 169 |
+
"loss": 3.9683,
|
| 170 |
+
"step": 7500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.13370101111389654,
|
| 174 |
+
"grad_norm": 5.36767578125,
|
| 175 |
+
"learning_rate": 4.368332350214882e-05,
|
| 176 |
+
"loss": 3.91,
|
| 177 |
+
"step": 8000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.13370101111389654,
|
| 181 |
+
"eval_loss": 3.774301528930664,
|
| 182 |
+
"eval_runtime": 34.7154,
|
| 183 |
+
"eval_samples_per_second": 128.531,
|
| 184 |
+
"eval_steps_per_second": 16.074,
|
| 185 |
+
"step": 8000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1420573243085151,
|
| 189 |
+
"grad_norm": 6.756371021270752,
|
| 190 |
+
"learning_rate": 4.326198702283644e-05,
|
| 191 |
+
"loss": 3.9031,
|
| 192 |
+
"step": 8500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1504136375031336,
|
| 196 |
+
"grad_norm": 5.863716125488281,
|
| 197 |
+
"learning_rate": 4.2840650543524055e-05,
|
| 198 |
+
"loss": 3.8962,
|
| 199 |
+
"step": 9000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1504136375031336,
|
| 203 |
+
"eval_loss": 3.7521703243255615,
|
| 204 |
+
"eval_runtime": 34.7163,
|
| 205 |
+
"eval_samples_per_second": 128.527,
|
| 206 |
+
"eval_steps_per_second": 16.073,
|
| 207 |
+
"step": 9000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.15876995069775215,
|
| 211 |
+
"grad_norm": 4.68450403213501,
|
| 212 |
+
"learning_rate": 4.241931406421168e-05,
|
| 213 |
+
"loss": 3.853,
|
| 214 |
+
"step": 9500
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.16712626389237067,
|
| 218 |
+
"grad_norm": 5.654048442840576,
|
| 219 |
+
"learning_rate": 4.199882025785793e-05,
|
| 220 |
+
"loss": 3.8244,
|
| 221 |
+
"step": 10000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.16712626389237067,
|
| 225 |
+
"eval_loss": 3.7112536430358887,
|
| 226 |
+
"eval_runtime": 34.6707,
|
| 227 |
+
"eval_samples_per_second": 128.697,
|
| 228 |
+
"eval_steps_per_second": 16.094,
|
| 229 |
+
"step": 10000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.17548257708698922,
|
| 233 |
+
"grad_norm": 4.9623284339904785,
|
| 234 |
+
"learning_rate": 4.157748377854555e-05,
|
| 235 |
+
"loss": 3.8098,
|
| 236 |
+
"step": 10500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.18383889028160774,
|
| 240 |
+
"grad_norm": 9.91655158996582,
|
| 241 |
+
"learning_rate": 4.115614729923317e-05,
|
| 242 |
+
"loss": 3.7687,
|
| 243 |
+
"step": 11000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.18383889028160774,
|
| 247 |
+
"eval_loss": 3.6787209510803223,
|
| 248 |
+
"eval_runtime": 34.653,
|
| 249 |
+
"eval_samples_per_second": 128.762,
|
| 250 |
+
"eval_steps_per_second": 16.103,
|
| 251 |
+
"step": 11000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.1921952034762263,
|
| 255 |
+
"grad_norm": 7.150275707244873,
|
| 256 |
+
"learning_rate": 4.0734810819920794e-05,
|
| 257 |
+
"loss": 3.8006,
|
| 258 |
+
"step": 11500
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.20055151667084484,
|
| 262 |
+
"grad_norm": 3.9780406951904297,
|
| 263 |
+
"learning_rate": 4.031347434060841e-05,
|
| 264 |
+
"loss": 3.7853,
|
| 265 |
+
"step": 12000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.20055151667084484,
|
| 269 |
+
"eval_loss": 3.655853748321533,
|
| 270 |
+
"eval_runtime": 34.6805,
|
| 271 |
+
"eval_samples_per_second": 128.66,
|
| 272 |
+
"eval_steps_per_second": 16.09,
|
| 273 |
+
"step": 12000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.20890782986546336,
|
| 277 |
+
"grad_norm": 7.393148422241211,
|
| 278 |
+
"learning_rate": 3.989298053425466e-05,
|
| 279 |
+
"loss": 3.8083,
|
| 280 |
+
"step": 12500
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.2172641430600819,
|
| 284 |
+
"grad_norm": 4.079512119293213,
|
| 285 |
+
"learning_rate": 3.947164405494228e-05,
|
| 286 |
+
"loss": 3.7215,
|
| 287 |
+
"step": 13000
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.2172641430600819,
|
| 291 |
+
"eval_loss": 3.6416282653808594,
|
| 292 |
+
"eval_runtime": 34.6971,
|
| 293 |
+
"eval_samples_per_second": 128.599,
|
| 294 |
+
"eval_steps_per_second": 16.082,
|
| 295 |
+
"step": 13000
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.22562045625470042,
|
| 299 |
+
"grad_norm": 3.7765800952911377,
|
| 300 |
+
"learning_rate": 3.90503075756299e-05,
|
| 301 |
+
"loss": 3.7131,
|
| 302 |
+
"step": 13500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.23397676944931897,
|
| 306 |
+
"grad_norm": 5.007580280303955,
|
| 307 |
+
"learning_rate": 3.8628971096317526e-05,
|
| 308 |
+
"loss": 3.7509,
|
| 309 |
+
"step": 14000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.23397676944931897,
|
| 313 |
+
"eval_loss": 3.6096222400665283,
|
| 314 |
+
"eval_runtime": 34.7053,
|
| 315 |
+
"eval_samples_per_second": 128.568,
|
| 316 |
+
"eval_steps_per_second": 16.078,
|
| 317 |
+
"step": 14000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.2423330826439375,
|
| 321 |
+
"grad_norm": 4.108855724334717,
|
| 322 |
+
"learning_rate": 3.820763461700515e-05,
|
| 323 |
+
"loss": 3.7315,
|
| 324 |
+
"step": 14500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.25068939583855604,
|
| 328 |
+
"grad_norm": 5.825819492340088,
|
| 329 |
+
"learning_rate": 3.7786298137692763e-05,
|
| 330 |
+
"loss": 3.7185,
|
| 331 |
+
"step": 15000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.25068939583855604,
|
| 335 |
+
"eval_loss": 3.5938196182250977,
|
| 336 |
+
"eval_runtime": 34.6849,
|
| 337 |
+
"eval_samples_per_second": 128.644,
|
| 338 |
+
"eval_steps_per_second": 16.088,
|
| 339 |
+
"step": 15000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2590457090331746,
|
| 343 |
+
"grad_norm": 5.564838409423828,
|
| 344 |
+
"learning_rate": 3.736496165838038e-05,
|
| 345 |
+
"loss": 3.7059,
|
| 346 |
+
"step": 15500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2674020222277931,
|
| 350 |
+
"grad_norm": 5.582629203796387,
|
| 351 |
+
"learning_rate": 3.6943625179068e-05,
|
| 352 |
+
"loss": 3.6318,
|
| 353 |
+
"step": 16000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2674020222277931,
|
| 357 |
+
"eval_loss": 3.5812368392944336,
|
| 358 |
+
"eval_runtime": 34.698,
|
| 359 |
+
"eval_samples_per_second": 128.595,
|
| 360 |
+
"eval_steps_per_second": 16.082,
|
| 361 |
+
"step": 16000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.2757583354224116,
|
| 365 |
+
"grad_norm": 7.805865287780762,
|
| 366 |
+
"learning_rate": 3.652313137271425e-05,
|
| 367 |
+
"loss": 3.6836,
|
| 368 |
+
"step": 16500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2841146486170302,
|
| 372 |
+
"grad_norm": 7.230532169342041,
|
| 373 |
+
"learning_rate": 3.61026375663605e-05,
|
| 374 |
+
"loss": 3.6965,
|
| 375 |
+
"step": 17000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.2841146486170302,
|
| 379 |
+
"eval_loss": 3.5589852333068848,
|
| 380 |
+
"eval_runtime": 34.7222,
|
| 381 |
+
"eval_samples_per_second": 128.506,
|
| 382 |
+
"eval_steps_per_second": 16.07,
|
| 383 |
+
"step": 17000
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.2924709618116487,
|
| 387 |
+
"grad_norm": 5.060893535614014,
|
| 388 |
+
"learning_rate": 3.568130108704812e-05,
|
| 389 |
+
"loss": 3.6219,
|
| 390 |
+
"step": 17500
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.3008272750062672,
|
| 394 |
+
"grad_norm": 6.640148639678955,
|
| 395 |
+
"learning_rate": 3.525996460773574e-05,
|
| 396 |
+
"loss": 3.6204,
|
| 397 |
+
"step": 18000
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.3008272750062672,
|
| 401 |
+
"eval_loss": 3.5444483757019043,
|
| 402 |
+
"eval_runtime": 34.8334,
|
| 403 |
+
"eval_samples_per_second": 128.095,
|
| 404 |
+
"eval_steps_per_second": 16.019,
|
| 405 |
+
"step": 18000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.30918358820088576,
|
| 409 |
+
"grad_norm": 6.7097930908203125,
|
| 410 |
+
"learning_rate": 3.483862812842336e-05,
|
| 411 |
+
"loss": 3.6134,
|
| 412 |
+
"step": 18500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.3175399013955043,
|
| 416 |
+
"grad_norm": 5.851566314697266,
|
| 417 |
+
"learning_rate": 3.441729164911098e-05,
|
| 418 |
+
"loss": 3.6138,
|
| 419 |
+
"step": 19000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.3175399013955043,
|
| 423 |
+
"eval_loss": 3.529630184173584,
|
| 424 |
+
"eval_runtime": 34.7073,
|
| 425 |
+
"eval_samples_per_second": 128.561,
|
| 426 |
+
"eval_steps_per_second": 16.077,
|
| 427 |
+
"step": 19000
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.32589621459012286,
|
| 431 |
+
"grad_norm": 4.4097981452941895,
|
| 432 |
+
"learning_rate": 3.3995955169798604e-05,
|
| 433 |
+
"loss": 3.6202,
|
| 434 |
+
"step": 19500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.33425252778474135,
|
| 438 |
+
"grad_norm": 6.881157398223877,
|
| 439 |
+
"learning_rate": 3.357461869048623e-05,
|
| 440 |
+
"loss": 3.5862,
|
| 441 |
+
"step": 20000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.33425252778474135,
|
| 445 |
+
"eval_loss": 3.5189144611358643,
|
| 446 |
+
"eval_runtime": 34.7046,
|
| 447 |
+
"eval_samples_per_second": 128.571,
|
| 448 |
+
"eval_steps_per_second": 16.079,
|
| 449 |
+
"step": 20000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.3426088409793599,
|
| 453 |
+
"grad_norm": 4.771751880645752,
|
| 454 |
+
"learning_rate": 3.315328221117385e-05,
|
| 455 |
+
"loss": 3.613,
|
| 456 |
+
"step": 20500
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.35096515417397844,
|
| 460 |
+
"grad_norm": 3.5520098209381104,
|
| 461 |
+
"learning_rate": 3.273278840482009e-05,
|
| 462 |
+
"loss": 3.5767,
|
| 463 |
+
"step": 21000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.35096515417397844,
|
| 467 |
+
"eval_loss": 3.506736993789673,
|
| 468 |
+
"eval_runtime": 35.4581,
|
| 469 |
+
"eval_samples_per_second": 125.839,
|
| 470 |
+
"eval_steps_per_second": 15.737,
|
| 471 |
+
"step": 21000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.359321467368597,
|
| 475 |
+
"grad_norm": 3.511641263961792,
|
| 476 |
+
"learning_rate": 3.2311451925507714e-05,
|
| 477 |
+
"loss": 3.5692,
|
| 478 |
+
"step": 21500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.3676777805632155,
|
| 482 |
+
"grad_norm": 4.536217212677002,
|
| 483 |
+
"learning_rate": 3.1890115446195336e-05,
|
| 484 |
+
"loss": 3.5913,
|
| 485 |
+
"step": 22000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.3676777805632155,
|
| 489 |
+
"eval_loss": 3.490816354751587,
|
| 490 |
+
"eval_runtime": 34.731,
|
| 491 |
+
"eval_samples_per_second": 128.473,
|
| 492 |
+
"eval_steps_per_second": 16.066,
|
| 493 |
+
"step": 22000
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.37603409375783403,
|
| 497 |
+
"grad_norm": 5.8858184814453125,
|
| 498 |
+
"learning_rate": 3.146877896688296e-05,
|
| 499 |
+
"loss": 3.6495,
|
| 500 |
+
"step": 22500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3843904069524526,
|
| 504 |
+
"grad_norm": 5.291027069091797,
|
| 505 |
+
"learning_rate": 3.10482851605292e-05,
|
| 506 |
+
"loss": 3.5796,
|
| 507 |
+
"step": 23000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.3843904069524526,
|
| 511 |
+
"eval_loss": 3.4832894802093506,
|
| 512 |
+
"eval_runtime": 34.6998,
|
| 513 |
+
"eval_samples_per_second": 128.589,
|
| 514 |
+
"eval_steps_per_second": 16.081,
|
| 515 |
+
"step": 23000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.3927467201470711,
|
| 519 |
+
"grad_norm": 4.761454105377197,
|
| 520 |
+
"learning_rate": 3.062694868121682e-05,
|
| 521 |
+
"loss": 3.5603,
|
| 522 |
+
"step": 23500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.4011030333416897,
|
| 526 |
+
"grad_norm": 4.90023946762085,
|
| 527 |
+
"learning_rate": 3.0205612201904442e-05,
|
| 528 |
+
"loss": 3.5428,
|
| 529 |
+
"step": 24000
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.4011030333416897,
|
| 533 |
+
"eval_loss": 3.4762682914733887,
|
| 534 |
+
"eval_runtime": 34.6953,
|
| 535 |
+
"eval_samples_per_second": 128.605,
|
| 536 |
+
"eval_steps_per_second": 16.083,
|
| 537 |
+
"step": 24000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.40945934653630817,
|
| 541 |
+
"grad_norm": 6.675970554351807,
|
| 542 |
+
"learning_rate": 2.9784275722592064e-05,
|
| 543 |
+
"loss": 3.5559,
|
| 544 |
+
"step": 24500
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.4178156597309267,
|
| 548 |
+
"grad_norm": 4.5930914878845215,
|
| 549 |
+
"learning_rate": 2.936378191623831e-05,
|
| 550 |
+
"loss": 3.5087,
|
| 551 |
+
"step": 25000
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.4178156597309267,
|
| 555 |
+
"eval_loss": 3.465832233428955,
|
| 556 |
+
"eval_runtime": 34.6863,
|
| 557 |
+
"eval_samples_per_second": 128.639,
|
| 558 |
+
"eval_steps_per_second": 16.087,
|
| 559 |
+
"step": 25000
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.42617197292554526,
|
| 563 |
+
"grad_norm": 4.835921287536621,
|
| 564 |
+
"learning_rate": 2.894244543692593e-05,
|
| 565 |
+
"loss": 3.5474,
|
| 566 |
+
"step": 25500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.4345282861201638,
|
| 570 |
+
"grad_norm": 5.828157424926758,
|
| 571 |
+
"learning_rate": 2.852110895761355e-05,
|
| 572 |
+
"loss": 3.5724,
|
| 573 |
+
"step": 26000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.4345282861201638,
|
| 577 |
+
"eval_loss": 3.458848714828491,
|
| 578 |
+
"eval_runtime": 34.6689,
|
| 579 |
+
"eval_samples_per_second": 128.703,
|
| 580 |
+
"eval_steps_per_second": 16.095,
|
| 581 |
+
"step": 26000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.4428845993147823,
|
| 585 |
+
"grad_norm": 3.7706058025360107,
|
| 586 |
+
"learning_rate": 2.8099772478301174e-05,
|
| 587 |
+
"loss": 3.49,
|
| 588 |
+
"step": 26500
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.45124091250940085,
|
| 592 |
+
"grad_norm": 4.761475563049316,
|
| 593 |
+
"learning_rate": 2.7679278671947416e-05,
|
| 594 |
+
"loss": 3.5493,
|
| 595 |
+
"step": 27000
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.45124091250940085,
|
| 599 |
+
"eval_loss": 3.447903633117676,
|
| 600 |
+
"eval_runtime": 34.6802,
|
| 601 |
+
"eval_samples_per_second": 128.661,
|
| 602 |
+
"eval_steps_per_second": 16.09,
|
| 603 |
+
"step": 27000
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.4595972257040194,
|
| 607 |
+
"grad_norm": 4.8344292640686035,
|
| 608 |
+
"learning_rate": 2.7258784865593662e-05,
|
| 609 |
+
"loss": 3.5161,
|
| 610 |
+
"step": 27500
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.46795353889863794,
|
| 614 |
+
"grad_norm": 8.372416496276855,
|
| 615 |
+
"learning_rate": 2.6837448386281284e-05,
|
| 616 |
+
"loss": 3.4814,
|
| 617 |
+
"step": 28000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.46795353889863794,
|
| 621 |
+
"eval_loss": 3.4423773288726807,
|
| 622 |
+
"eval_runtime": 34.6998,
|
| 623 |
+
"eval_samples_per_second": 128.588,
|
| 624 |
+
"eval_steps_per_second": 16.081,
|
| 625 |
+
"step": 28000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.47630985209325644,
|
| 629 |
+
"grad_norm": 3.9577817916870117,
|
| 630 |
+
"learning_rate": 2.6416111906968903e-05,
|
| 631 |
+
"loss": 3.5169,
|
| 632 |
+
"step": 28500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.484666165287875,
|
| 636 |
+
"grad_norm": 4.661665439605713,
|
| 637 |
+
"learning_rate": 2.5994775427656526e-05,
|
| 638 |
+
"loss": 3.5444,
|
| 639 |
+
"step": 29000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.484666165287875,
|
| 643 |
+
"eval_loss": 3.4300851821899414,
|
| 644 |
+
"eval_runtime": 34.6929,
|
| 645 |
+
"eval_samples_per_second": 128.614,
|
| 646 |
+
"eval_steps_per_second": 16.084,
|
| 647 |
+
"step": 29000
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.49302247848249353,
|
| 651 |
+
"grad_norm": 5.853386402130127,
|
| 652 |
+
"learning_rate": 2.5573438948344148e-05,
|
| 653 |
+
"loss": 3.5049,
|
| 654 |
+
"step": 29500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5013787916771121,
|
| 658 |
+
"grad_norm": 3.900343894958496,
|
| 659 |
+
"learning_rate": 2.515210246903177e-05,
|
| 660 |
+
"loss": 3.5451,
|
| 661 |
+
"step": 30000
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5013787916771121,
|
| 665 |
+
"eval_loss": 3.421967029571533,
|
| 666 |
+
"eval_runtime": 34.6635,
|
| 667 |
+
"eval_samples_per_second": 128.723,
|
| 668 |
+
"eval_steps_per_second": 16.098,
|
| 669 |
+
"step": 30000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.5097351048717306,
|
| 673 |
+
"grad_norm": 5.265810966491699,
|
| 674 |
+
"learning_rate": 2.473076598971939e-05,
|
| 675 |
+
"loss": 3.4952,
|
| 676 |
+
"step": 30500
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.5180914180663492,
|
| 680 |
+
"grad_norm": 6.181280612945557,
|
| 681 |
+
"learning_rate": 2.430942951040701e-05,
|
| 682 |
+
"loss": 3.4879,
|
| 683 |
+
"step": 31000
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.5180914180663492,
|
| 687 |
+
"eval_loss": 3.4192793369293213,
|
| 688 |
+
"eval_runtime": 34.6959,
|
| 689 |
+
"eval_samples_per_second": 128.603,
|
| 690 |
+
"eval_steps_per_second": 16.083,
|
| 691 |
+
"step": 31000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.5264477312609677,
|
| 695 |
+
"grad_norm": 7.2022294998168945,
|
| 696 |
+
"learning_rate": 2.3888093031094634e-05,
|
| 697 |
+
"loss": 3.5258,
|
| 698 |
+
"step": 31500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.5348040444555862,
|
| 702 |
+
"grad_norm": 4.311310291290283,
|
| 703 |
+
"learning_rate": 2.3468441897699503e-05,
|
| 704 |
+
"loss": 3.4957,
|
| 705 |
+
"step": 32000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.5348040444555862,
|
| 709 |
+
"eval_loss": 3.406111240386963,
|
| 710 |
+
"eval_runtime": 34.7236,
|
| 711 |
+
"eval_samples_per_second": 128.501,
|
| 712 |
+
"eval_steps_per_second": 16.07,
|
| 713 |
+
"step": 32000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.5431603576502048,
|
| 717 |
+
"grad_norm": 4.1522722244262695,
|
| 718 |
+
"learning_rate": 2.3047105418387125e-05,
|
| 719 |
+
"loss": 3.5138,
|
| 720 |
+
"step": 32500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.5515166708448233,
|
| 724 |
+
"grad_norm": 4.656116962432861,
|
| 725 |
+
"learning_rate": 2.2625768939074744e-05,
|
| 726 |
+
"loss": 3.4637,
|
| 727 |
+
"step": 33000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.5515166708448233,
|
| 731 |
+
"eval_loss": 3.3979318141937256,
|
| 732 |
+
"eval_runtime": 34.7041,
|
| 733 |
+
"eval_samples_per_second": 128.573,
|
| 734 |
+
"eval_steps_per_second": 16.079,
|
| 735 |
+
"step": 33000
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.5598729840394417,
|
| 739 |
+
"grad_norm": 7.333806037902832,
|
| 740 |
+
"learning_rate": 2.2204432459762367e-05,
|
| 741 |
+
"loss": 3.5071,
|
| 742 |
+
"step": 33500
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.5682292972340603,
|
| 746 |
+
"grad_norm": 4.071677207946777,
|
| 747 |
+
"learning_rate": 2.178309598044999e-05,
|
| 748 |
+
"loss": 3.4836,
|
| 749 |
+
"step": 34000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.5682292972340603,
|
| 753 |
+
"eval_loss": 3.396083354949951,
|
| 754 |
+
"eval_runtime": 34.7068,
|
| 755 |
+
"eval_samples_per_second": 128.563,
|
| 756 |
+
"eval_steps_per_second": 16.078,
|
| 757 |
+
"step": 34000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.5765856104286788,
|
| 761 |
+
"grad_norm": 4.475882053375244,
|
| 762 |
+
"learning_rate": 2.136175950113761e-05,
|
| 763 |
+
"loss": 3.4764,
|
| 764 |
+
"step": 34500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.5849419236232974,
|
| 768 |
+
"grad_norm": 3.323747158050537,
|
| 769 |
+
"learning_rate": 2.094042302182523e-05,
|
| 770 |
+
"loss": 3.4379,
|
| 771 |
+
"step": 35000
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.5849419236232974,
|
| 775 |
+
"eval_loss": 3.389554977416992,
|
| 776 |
+
"eval_runtime": 34.7078,
|
| 777 |
+
"eval_samples_per_second": 128.559,
|
| 778 |
+
"eval_steps_per_second": 16.077,
|
| 779 |
+
"step": 35000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5932982368179159,
|
| 783 |
+
"grad_norm": 6.1050543785095215,
|
| 784 |
+
"learning_rate": 2.0519086542512852e-05,
|
| 785 |
+
"loss": 3.4959,
|
| 786 |
+
"step": 35500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.6016545500125344,
|
| 790 |
+
"grad_norm": 5.739749431610107,
|
| 791 |
+
"learning_rate": 2.0097750063200475e-05,
|
| 792 |
+
"loss": 3.4672,
|
| 793 |
+
"step": 36000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6016545500125344,
|
| 797 |
+
"eval_loss": 3.3847856521606445,
|
| 798 |
+
"eval_runtime": 34.7286,
|
| 799 |
+
"eval_samples_per_second": 128.482,
|
| 800 |
+
"eval_steps_per_second": 16.067,
|
| 801 |
+
"step": 36000
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.610010863207153,
|
| 805 |
+
"grad_norm": 4.3374924659729,
|
| 806 |
+
"learning_rate": 1.9677256256846717e-05,
|
| 807 |
+
"loss": 3.4903,
|
| 808 |
+
"step": 36500
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6183671764017715,
|
| 812 |
+
"grad_norm": 5.808351516723633,
|
| 813 |
+
"learning_rate": 1.925591977753434e-05,
|
| 814 |
+
"loss": 3.4723,
|
| 815 |
+
"step": 37000
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.6183671764017715,
|
| 819 |
+
"eval_loss": 3.379159450531006,
|
| 820 |
+
"eval_runtime": 34.676,
|
| 821 |
+
"eval_samples_per_second": 128.677,
|
| 822 |
+
"eval_steps_per_second": 16.092,
|
| 823 |
+
"step": 37000
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.6267234895963901,
|
| 827 |
+
"grad_norm": 4.286294460296631,
|
| 828 |
+
"learning_rate": 1.8834583298221962e-05,
|
| 829 |
+
"loss": 3.4922,
|
| 830 |
+
"step": 37500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.6350798027910086,
|
| 834 |
+
"grad_norm": 4.990993976593018,
|
| 835 |
+
"learning_rate": 1.8413246818909584e-05,
|
| 836 |
+
"loss": 3.4815,
|
| 837 |
+
"step": 38000
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.6350798027910086,
|
| 841 |
+
"eval_loss": 3.376216411590576,
|
| 842 |
+
"eval_runtime": 34.7129,
|
| 843 |
+
"eval_samples_per_second": 128.54,
|
| 844 |
+
"eval_steps_per_second": 16.075,
|
| 845 |
+
"step": 38000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.6434361159856271,
|
| 849 |
+
"grad_norm": 4.532817840576172,
|
| 850 |
+
"learning_rate": 1.7992753012555827e-05,
|
| 851 |
+
"loss": 3.4146,
|
| 852 |
+
"step": 38500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.6517924291802457,
|
| 856 |
+
"grad_norm": 5.179249286651611,
|
| 857 |
+
"learning_rate": 1.757141653324345e-05,
|
| 858 |
+
"loss": 3.4357,
|
| 859 |
+
"step": 39000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.6517924291802457,
|
| 863 |
+
"eval_loss": 3.3684535026550293,
|
| 864 |
+
"eval_runtime": 34.5425,
|
| 865 |
+
"eval_samples_per_second": 129.174,
|
| 866 |
+
"eval_steps_per_second": 16.154,
|
| 867 |
+
"step": 39000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.6601487423748642,
|
| 871 |
+
"grad_norm": 5.6737542152404785,
|
| 872 |
+
"learning_rate": 1.715008005393107e-05,
|
| 873 |
+
"loss": 3.3936,
|
| 874 |
+
"step": 39500
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.6685050555694827,
|
| 878 |
+
"grad_norm": 4.707485675811768,
|
| 879 |
+
"learning_rate": 1.672874357461869e-05,
|
| 880 |
+
"loss": 3.4711,
|
| 881 |
+
"step": 40000
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.6685050555694827,
|
| 885 |
+
"eval_loss": 3.3646059036254883,
|
| 886 |
+
"eval_runtime": 34.5756,
|
| 887 |
+
"eval_samples_per_second": 129.051,
|
| 888 |
+
"eval_steps_per_second": 16.139,
|
| 889 |
+
"step": 40000
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.6768613687641013,
|
| 893 |
+
"grad_norm": 4.501412868499756,
|
| 894 |
+
"learning_rate": 1.630824976826494e-05,
|
| 895 |
+
"loss": 3.4073,
|
| 896 |
+
"step": 40500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.6852176819587198,
|
| 900 |
+
"grad_norm": 9.65688419342041,
|
| 901 |
+
"learning_rate": 1.5886913288952558e-05,
|
| 902 |
+
"loss": 3.4437,
|
| 903 |
+
"step": 41000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.6852176819587198,
|
| 907 |
+
"eval_loss": 3.3600456714630127,
|
| 908 |
+
"eval_runtime": 34.9587,
|
| 909 |
+
"eval_samples_per_second": 127.636,
|
| 910 |
+
"eval_steps_per_second": 15.962,
|
| 911 |
+
"step": 41000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.6935739951533384,
|
| 915 |
+
"grad_norm": 3.365447998046875,
|
| 916 |
+
"learning_rate": 1.5465576809640177e-05,
|
| 917 |
+
"loss": 3.3704,
|
| 918 |
+
"step": 41500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.7019303083479569,
|
| 922 |
+
"grad_norm": 4.373350143432617,
|
| 923 |
+
"learning_rate": 1.50442403303278e-05,
|
| 924 |
+
"loss": 3.4171,
|
| 925 |
+
"step": 42000
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.7019303083479569,
|
| 929 |
+
"eval_loss": 3.3585731983184814,
|
| 930 |
+
"eval_runtime": 34.556,
|
| 931 |
+
"eval_samples_per_second": 129.124,
|
| 932 |
+
"eval_steps_per_second": 16.148,
|
| 933 |
+
"step": 42000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.7102866215425754,
|
| 937 |
+
"grad_norm": 6.150047779083252,
|
| 938 |
+
"learning_rate": 1.4623746523974047e-05,
|
| 939 |
+
"loss": 3.3822,
|
| 940 |
+
"step": 42500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.718642934737194,
|
| 944 |
+
"grad_norm": 5.976407527923584,
|
| 945 |
+
"learning_rate": 1.4202410044661668e-05,
|
| 946 |
+
"loss": 3.442,
|
| 947 |
+
"step": 43000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.718642934737194,
|
| 951 |
+
"eval_loss": 3.3521482944488525,
|
| 952 |
+
"eval_runtime": 34.7345,
|
| 953 |
+
"eval_samples_per_second": 128.46,
|
| 954 |
+
"eval_steps_per_second": 16.065,
|
| 955 |
+
"step": 43000
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.7269992479318125,
|
| 959 |
+
"grad_norm": 5.317568778991699,
|
| 960 |
+
"learning_rate": 1.378107356534929e-05,
|
| 961 |
+
"loss": 3.3837,
|
| 962 |
+
"step": 43500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.735355561126431,
|
| 966 |
+
"grad_norm": 4.642706871032715,
|
| 967 |
+
"learning_rate": 1.335973708603691e-05,
|
| 968 |
+
"loss": 3.3847,
|
| 969 |
+
"step": 44000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.735355561126431,
|
| 973 |
+
"eval_loss": 3.3473587036132812,
|
| 974 |
+
"eval_runtime": 34.8742,
|
| 975 |
+
"eval_samples_per_second": 127.946,
|
| 976 |
+
"eval_steps_per_second": 16.0,
|
| 977 |
+
"step": 44000
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.7437118743210496,
|
| 981 |
+
"grad_norm": 7.499514102935791,
|
| 982 |
+
"learning_rate": 1.293840060672453e-05,
|
| 983 |
+
"loss": 3.423,
|
| 984 |
+
"step": 44500
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.7520681875156681,
|
| 988 |
+
"grad_norm": 4.7019548416137695,
|
| 989 |
+
"learning_rate": 1.2517906800370777e-05,
|
| 990 |
+
"loss": 3.4064,
|
| 991 |
+
"step": 45000
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.7520681875156681,
|
| 995 |
+
"eval_loss": 3.347837209701538,
|
| 996 |
+
"eval_runtime": 34.5258,
|
| 997 |
+
"eval_samples_per_second": 129.237,
|
| 998 |
+
"eval_steps_per_second": 16.162,
|
| 999 |
+
"step": 45000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.7604245007102867,
|
| 1003 |
+
"grad_norm": 6.7862725257873535,
|
| 1004 |
+
"learning_rate": 1.2096570321058398e-05,
|
| 1005 |
+
"loss": 3.4665,
|
| 1006 |
+
"step": 45500
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.7687808139049052,
|
| 1010 |
+
"grad_norm": 7.120584964752197,
|
| 1011 |
+
"learning_rate": 1.167523384174602e-05,
|
| 1012 |
+
"loss": 3.4181,
|
| 1013 |
+
"step": 46000
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.7687808139049052,
|
| 1017 |
+
"eval_loss": 3.3389999866485596,
|
| 1018 |
+
"eval_runtime": 34.5968,
|
| 1019 |
+
"eval_samples_per_second": 128.972,
|
| 1020 |
+
"eval_steps_per_second": 16.129,
|
| 1021 |
+
"step": 46000
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 0.7771371270995237,
|
| 1025 |
+
"grad_norm": 6.571496963500977,
|
| 1026 |
+
"learning_rate": 1.1253897362433639e-05,
|
| 1027 |
+
"loss": 3.3691,
|
| 1028 |
+
"step": 46500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.7854934402941423,
|
| 1032 |
+
"grad_norm": 4.248281955718994,
|
| 1033 |
+
"learning_rate": 1.0833403556079886e-05,
|
| 1034 |
+
"loss": 3.3822,
|
| 1035 |
+
"step": 47000
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.7854934402941423,
|
| 1039 |
+
"eval_loss": 3.334426164627075,
|
| 1040 |
+
"eval_runtime": 34.5427,
|
| 1041 |
+
"eval_samples_per_second": 129.173,
|
| 1042 |
+
"eval_steps_per_second": 16.154,
|
| 1043 |
+
"step": 47000
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.7938497534887607,
|
| 1047 |
+
"grad_norm": 8.167221069335938,
|
| 1048 |
+
"learning_rate": 1.0412067076767507e-05,
|
| 1049 |
+
"loss": 3.4073,
|
| 1050 |
+
"step": 47500
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.8022060666833793,
|
| 1054 |
+
"grad_norm": 4.705054759979248,
|
| 1055 |
+
"learning_rate": 9.990730597455128e-06,
|
| 1056 |
+
"loss": 3.4277,
|
| 1057 |
+
"step": 48000
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.8022060666833793,
|
| 1061 |
+
"eval_loss": 3.3343007564544678,
|
| 1062 |
+
"eval_runtime": 34.5199,
|
| 1063 |
+
"eval_samples_per_second": 129.259,
|
| 1064 |
+
"eval_steps_per_second": 16.165,
|
| 1065 |
+
"step": 48000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.8105623798779978,
|
| 1069 |
+
"grad_norm": 5.092607498168945,
|
| 1070 |
+
"learning_rate": 9.56939411814275e-06,
|
| 1071 |
+
"loss": 3.4014,
|
| 1072 |
+
"step": 48500
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.8189186930726163,
|
| 1076 |
+
"grad_norm": 4.517687797546387,
|
| 1077 |
+
"learning_rate": 9.148900311788996e-06,
|
| 1078 |
+
"loss": 3.331,
|
| 1079 |
+
"step": 49000
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.8189186930726163,
|
| 1083 |
+
"eval_loss": 3.332448720932007,
|
| 1084 |
+
"eval_runtime": 34.966,
|
| 1085 |
+
"eval_samples_per_second": 127.61,
|
| 1086 |
+
"eval_steps_per_second": 15.958,
|
| 1087 |
+
"step": 49000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.8272750062672349,
|
| 1091 |
+
"grad_norm": 5.909774303436279,
|
| 1092 |
+
"learning_rate": 8.727563832476616e-06,
|
| 1093 |
+
"loss": 3.4098,
|
| 1094 |
+
"step": 49500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.8356313194618534,
|
| 1098 |
+
"grad_norm": 4.237793922424316,
|
| 1099 |
+
"learning_rate": 8.306227353164239e-06,
|
| 1100 |
+
"loss": 3.4073,
|
| 1101 |
+
"step": 50000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.8356313194618534,
|
| 1105 |
+
"eval_loss": 3.328007459640503,
|
| 1106 |
+
"eval_runtime": 35.0133,
|
| 1107 |
+
"eval_samples_per_second": 127.437,
|
| 1108 |
+
"eval_steps_per_second": 15.937,
|
| 1109 |
+
"step": 50000
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.8439876326564719,
|
| 1113 |
+
"grad_norm": 6.143158912658691,
|
| 1114 |
+
"learning_rate": 7.884890873851858e-06,
|
| 1115 |
+
"loss": 3.3904,
|
| 1116 |
+
"step": 50500
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.8523439458510905,
|
| 1120 |
+
"grad_norm": 7.914546966552734,
|
| 1121 |
+
"learning_rate": 7.464397067498103e-06,
|
| 1122 |
+
"loss": 3.3741,
|
| 1123 |
+
"step": 51000
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.8523439458510905,
|
| 1127 |
+
"eval_loss": 3.327684164047241,
|
| 1128 |
+
"eval_runtime": 34.6263,
|
| 1129 |
+
"eval_samples_per_second": 128.862,
|
| 1130 |
+
"eval_steps_per_second": 16.115,
|
| 1131 |
+
"step": 51000
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.860700259045709,
|
| 1135 |
+
"grad_norm": 4.859989643096924,
|
| 1136 |
+
"learning_rate": 7.043060588185725e-06,
|
| 1137 |
+
"loss": 3.3614,
|
| 1138 |
+
"step": 51500
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.8690565722403276,
|
| 1142 |
+
"grad_norm": 4.473916053771973,
|
| 1143 |
+
"learning_rate": 6.621724108873346e-06,
|
| 1144 |
+
"loss": 3.3467,
|
| 1145 |
+
"step": 52000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.8690565722403276,
|
| 1149 |
+
"eval_loss": 3.3206727504730225,
|
| 1150 |
+
"eval_runtime": 34.5416,
|
| 1151 |
+
"eval_samples_per_second": 129.178,
|
| 1152 |
+
"eval_steps_per_second": 16.154,
|
| 1153 |
+
"step": 52000
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.8774128854349461,
|
| 1157 |
+
"grad_norm": 5.406485080718994,
|
| 1158 |
+
"learning_rate": 6.200387629560968e-06,
|
| 1159 |
+
"loss": 3.3221,
|
| 1160 |
+
"step": 52500
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.8857691986295646,
|
| 1164 |
+
"grad_norm": 5.721664905548096,
|
| 1165 |
+
"learning_rate": 5.7798938232072135e-06,
|
| 1166 |
+
"loss": 3.418,
|
| 1167 |
+
"step": 53000
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.8857691986295646,
|
| 1171 |
+
"eval_loss": 3.320549488067627,
|
| 1172 |
+
"eval_runtime": 34.5125,
|
| 1173 |
+
"eval_samples_per_second": 129.287,
|
| 1174 |
+
"eval_steps_per_second": 16.168,
|
| 1175 |
+
"step": 53000
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.8941255118241832,
|
| 1179 |
+
"grad_norm": 8.336995124816895,
|
| 1180 |
+
"learning_rate": 5.358557343894835e-06,
|
| 1181 |
+
"loss": 3.3746,
|
| 1182 |
+
"step": 53500
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.9024818250188017,
|
| 1186 |
+
"grad_norm": 6.922939777374268,
|
| 1187 |
+
"learning_rate": 4.937220864582456e-06,
|
| 1188 |
+
"loss": 3.416,
|
| 1189 |
+
"step": 54000
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.9024818250188017,
|
| 1193 |
+
"eval_loss": 3.3170385360717773,
|
| 1194 |
+
"eval_runtime": 34.5291,
|
| 1195 |
+
"eval_samples_per_second": 129.224,
|
| 1196 |
+
"eval_steps_per_second": 16.16,
|
| 1197 |
+
"step": 54000
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.9108381382134202,
|
| 1201 |
+
"grad_norm": 6.568256855010986,
|
| 1202 |
+
"learning_rate": 4.515884385270077e-06,
|
| 1203 |
+
"loss": 3.3983,
|
| 1204 |
+
"step": 54500
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 0.9191944514080388,
|
| 1208 |
+
"grad_norm": 4.785107612609863,
|
| 1209 |
+
"learning_rate": 4.095390578916323e-06,
|
| 1210 |
+
"loss": 3.3607,
|
| 1211 |
+
"step": 55000
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.9191944514080388,
|
| 1215 |
+
"eval_loss": 3.3168444633483887,
|
| 1216 |
+
"eval_runtime": 34.6371,
|
| 1217 |
+
"eval_samples_per_second": 128.821,
|
| 1218 |
+
"eval_steps_per_second": 16.11,
|
| 1219 |
+
"step": 55000
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 0.9275507646026573,
|
| 1223 |
+
"grad_norm": 5.468000888824463,
|
| 1224 |
+
"learning_rate": 3.6740540996039444e-06,
|
| 1225 |
+
"loss": 3.3587,
|
| 1226 |
+
"step": 55500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.9359070777972759,
|
| 1230 |
+
"grad_norm": 8.031869888305664,
|
| 1231 |
+
"learning_rate": 3.252717620291565e-06,
|
| 1232 |
+
"loss": 3.4075,
|
| 1233 |
+
"step": 56000
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.9359070777972759,
|
| 1237 |
+
"eval_loss": 3.3138468265533447,
|
| 1238 |
+
"eval_runtime": 34.5147,
|
| 1239 |
+
"eval_samples_per_second": 129.278,
|
| 1240 |
+
"eval_steps_per_second": 16.167,
|
| 1241 |
+
"step": 56000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.9442633909918944,
|
| 1245 |
+
"grad_norm": 5.647308349609375,
|
| 1246 |
+
"learning_rate": 2.831381140979186e-06,
|
| 1247 |
+
"loss": 3.4308,
|
| 1248 |
+
"step": 56500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.9526197041865129,
|
| 1252 |
+
"grad_norm": 3.991074562072754,
|
| 1253 |
+
"learning_rate": 2.410044661666807e-06,
|
| 1254 |
+
"loss": 3.3257,
|
| 1255 |
+
"step": 57000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.9526197041865129,
|
| 1259 |
+
"eval_loss": 3.3135740756988525,
|
| 1260 |
+
"eval_runtime": 34.9614,
|
| 1261 |
+
"eval_samples_per_second": 127.627,
|
| 1262 |
+
"eval_steps_per_second": 15.96,
|
| 1263 |
+
"step": 57000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.9609760173811315,
|
| 1267 |
+
"grad_norm": 4.273293972015381,
|
| 1268 |
+
"learning_rate": 1.9895508553130533e-06,
|
| 1269 |
+
"loss": 3.3362,
|
| 1270 |
+
"step": 57500
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.96933233057575,
|
| 1274 |
+
"grad_norm": 4.560577869415283,
|
| 1275 |
+
"learning_rate": 1.5682143760006742e-06,
|
| 1276 |
+
"loss": 3.3657,
|
| 1277 |
+
"step": 58000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.96933233057575,
|
| 1281 |
+
"eval_loss": 3.312079668045044,
|
| 1282 |
+
"eval_runtime": 34.557,
|
| 1283 |
+
"eval_samples_per_second": 129.12,
|
| 1284 |
+
"eval_steps_per_second": 16.147,
|
| 1285 |
+
"step": 58000
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.9776886437703685,
|
| 1289 |
+
"grad_norm": 3.613813877105713,
|
| 1290 |
+
"learning_rate": 1.1468778966882954e-06,
|
| 1291 |
+
"loss": 3.3847,
|
| 1292 |
+
"step": 58500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.9860449569649871,
|
| 1296 |
+
"grad_norm": 5.713568687438965,
|
| 1297 |
+
"learning_rate": 7.255414173759165e-07,
|
| 1298 |
+
"loss": 3.3508,
|
| 1299 |
+
"step": 59000
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.9860449569649871,
|
| 1303 |
+
"eval_loss": 3.311403751373291,
|
| 1304 |
+
"eval_runtime": 34.5443,
|
| 1305 |
+
"eval_samples_per_second": 129.167,
|
| 1306 |
+
"eval_steps_per_second": 16.153,
|
| 1307 |
+
"step": 59000
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.9944012701596056,
|
| 1311 |
+
"grad_norm": 8.699734687805176,
|
| 1312 |
+
"learning_rate": 3.050476110221623e-07,
|
| 1313 |
+
"loss": 3.3511,
|
| 1314 |
+
"step": 59500
|
| 1315 |
+
}
|
| 1316 |
+
],
|
| 1317 |
+
"logging_steps": 500,
|
| 1318 |
+
"max_steps": 59835,
|
| 1319 |
+
"num_input_tokens_seen": 0,
|
| 1320 |
+
"num_train_epochs": 1,
|
| 1321 |
+
"save_steps": 2000,
|
| 1322 |
+
"stateful_callbacks": {
|
| 1323 |
+
"TrainerControl": {
|
| 1324 |
+
"args": {
|
| 1325 |
+
"should_epoch_stop": false,
|
| 1326 |
+
"should_evaluate": false,
|
| 1327 |
+
"should_log": false,
|
| 1328 |
+
"should_save": true,
|
| 1329 |
+
"should_training_stop": true
|
| 1330 |
+
},
|
| 1331 |
+
"attributes": {}
|
| 1332 |
+
}
|
| 1333 |
+
},
|
| 1334 |
+
"total_flos": 2.50149494587392e+17,
|
| 1335 |
+
"train_batch_size": 8,
|
| 1336 |
+
"trial_name": null,
|
| 1337 |
+
"trial_params": null
|
| 1338 |
+
}
|
custom-tokenizer/full-finetuning/german-gpt2/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c53c27a55ef1914ee358ac472e9b6d5dfa998d6ba639f176cdb77cc70749b1b
|
| 3 |
+
size 5368
|
full-finetuning/LLaMmlein_120M/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "LSX-UniWue/LLaMmlein_120M",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 1,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"head_dim": 64,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 2048,
|
| 15 |
+
"max_position_embeddings": 2048,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"num_key_value_heads": 4,
|
| 21 |
+
"pretraining_tp": 1,
|
| 22 |
+
"rms_norm_eps": 1e-05,
|
| 23 |
+
"rope_scaling": null,
|
| 24 |
+
"rope_theta": 10000.0,
|
| 25 |
+
"tie_word_embeddings": false,
|
| 26 |
+
"torch_dtype": "float32",
|
| 27 |
+
"transformers_version": "4.46.0",
|
| 28 |
+
"use_cache": true,
|
| 29 |
+
"vocab_size": 32000
|
| 30 |
+
}
|
full-finetuning/LLaMmlein_120M/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.46.0"
|
| 6 |
+
}
|
full-finetuning/LLaMmlein_120M/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0399b28b6f631f75d922b415a46df1e1af6909b40028ed196413fb8b030e4fbe
|
| 3 |
+
size 498687008
|
full-finetuning/LLaMmlein_120M/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0896cf93d0e1bc98805dfc37179adcf8fe7d846d23741e64d13ac475dbc4626c
|
| 3 |
+
size 997443194
|
full-finetuning/LLaMmlein_120M/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3bc1ac846859ad13fd356d3ed41fd20834205a7e8764e4198636863fca3f64c
|
| 3 |
+
size 14244
|
full-finetuning/LLaMmlein_120M/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3844cc778b6d10aae1d91adae178f842d6cf04cca5898f583c90f0d4ec34a05
|
| 3 |
+
size 1064
|
full-finetuning/LLaMmlein_120M/trainer_state.json
ADDED
|
@@ -0,0 +1,1338 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.1603705883026123,
|
| 3 |
+
"best_model_checkpoint": "./models/full-finetuning/LLaMmlein_120M/checkpoint-58000",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 59835,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008356313194618534,
|
| 13 |
+
"grad_norm": 40.14425277709961,
|
| 14 |
+
"learning_rate": 4.96e-05,
|
| 15 |
+
"loss": 4.8477,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016712626389237067,
|
| 20 |
+
"grad_norm": 22.107275009155273,
|
| 21 |
+
"learning_rate": 4.958203421252212e-05,
|
| 22 |
+
"loss": 4.4466,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.016712626389237067,
|
| 27 |
+
"eval_loss": 4.3118462562561035,
|
| 28 |
+
"eval_runtime": 22.0404,
|
| 29 |
+
"eval_samples_per_second": 202.446,
|
| 30 |
+
"eval_steps_per_second": 25.317,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.025068939583855605,
|
| 35 |
+
"grad_norm": 19.609399795532227,
|
| 36 |
+
"learning_rate": 4.916069773320974e-05,
|
| 37 |
+
"loss": 4.2986,
|
| 38 |
+
"step": 1500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.033425252778474135,
|
| 42 |
+
"grad_norm": 18.64438247680664,
|
| 43 |
+
"learning_rate": 4.873936125389736e-05,
|
| 44 |
+
"loss": 4.2295,
|
| 45 |
+
"step": 2000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.033425252778474135,
|
| 49 |
+
"eval_loss": 4.121812343597412,
|
| 50 |
+
"eval_runtime": 22.1428,
|
| 51 |
+
"eval_samples_per_second": 201.51,
|
| 52 |
+
"eval_steps_per_second": 25.2,
|
| 53 |
+
"step": 2000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.04178156597309267,
|
| 57 |
+
"grad_norm": 19.098369598388672,
|
| 58 |
+
"learning_rate": 4.8318024774584986e-05,
|
| 59 |
+
"loss": 4.1165,
|
| 60 |
+
"step": 2500
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.05013787916771121,
|
| 64 |
+
"grad_norm": 17.83785629272461,
|
| 65 |
+
"learning_rate": 4.789668829527261e-05,
|
| 66 |
+
"loss": 4.029,
|
| 67 |
+
"step": 3000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.05013787916771121,
|
| 71 |
+
"eval_loss": 4.040452480316162,
|
| 72 |
+
"eval_runtime": 22.088,
|
| 73 |
+
"eval_samples_per_second": 202.01,
|
| 74 |
+
"eval_steps_per_second": 25.263,
|
| 75 |
+
"step": 3000
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.05849419236232974,
|
| 79 |
+
"grad_norm": 14.394288063049316,
|
| 80 |
+
"learning_rate": 4.747535181596023e-05,
|
| 81 |
+
"loss": 4.0701,
|
| 82 |
+
"step": 3500
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.06685050555694827,
|
| 86 |
+
"grad_norm": 17.005945205688477,
|
| 87 |
+
"learning_rate": 4.705401533664785e-05,
|
| 88 |
+
"loss": 4.0239,
|
| 89 |
+
"step": 4000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.06685050555694827,
|
| 93 |
+
"eval_loss": 3.948943853378296,
|
| 94 |
+
"eval_runtime": 22.0828,
|
| 95 |
+
"eval_samples_per_second": 202.058,
|
| 96 |
+
"eval_steps_per_second": 25.269,
|
| 97 |
+
"step": 4000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.0752068187515668,
|
| 101 |
+
"grad_norm": 20.297489166259766,
|
| 102 |
+
"learning_rate": 4.663267885733547e-05,
|
| 103 |
+
"loss": 3.9704,
|
| 104 |
+
"step": 4500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.08356313194618534,
|
| 108 |
+
"grad_norm": 17.280521392822266,
|
| 109 |
+
"learning_rate": 4.621134237802309e-05,
|
| 110 |
+
"loss": 3.9118,
|
| 111 |
+
"step": 5000
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.08356313194618534,
|
| 115 |
+
"eval_loss": 3.891711950302124,
|
| 116 |
+
"eval_runtime": 22.1454,
|
| 117 |
+
"eval_samples_per_second": 201.487,
|
| 118 |
+
"eval_steps_per_second": 25.197,
|
| 119 |
+
"step": 5000
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.09191944514080387,
|
| 123 |
+
"grad_norm": 18.847597122192383,
|
| 124 |
+
"learning_rate": 4.579000589871071e-05,
|
| 125 |
+
"loss": 3.9024,
|
| 126 |
+
"step": 5500
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.10027575833542242,
|
| 130 |
+
"grad_norm": 14.654472351074219,
|
| 131 |
+
"learning_rate": 4.5368669419398335e-05,
|
| 132 |
+
"loss": 3.8479,
|
| 133 |
+
"step": 6000
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.10027575833542242,
|
| 137 |
+
"eval_loss": 3.8378305435180664,
|
| 138 |
+
"eval_runtime": 22.1051,
|
| 139 |
+
"eval_samples_per_second": 201.854,
|
| 140 |
+
"eval_steps_per_second": 25.243,
|
| 141 |
+
"step": 6000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.10863207153004095,
|
| 145 |
+
"grad_norm": 18.267274856567383,
|
| 146 |
+
"learning_rate": 4.494817561304458e-05,
|
| 147 |
+
"loss": 3.8116,
|
| 148 |
+
"step": 6500
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.11698838472465949,
|
| 152 |
+
"grad_norm": 13.313653945922852,
|
| 153 |
+
"learning_rate": 4.45268391337322e-05,
|
| 154 |
+
"loss": 3.8592,
|
| 155 |
+
"step": 7000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.11698838472465949,
|
| 159 |
+
"eval_loss": 3.8072171211242676,
|
| 160 |
+
"eval_runtime": 22.2351,
|
| 161 |
+
"eval_samples_per_second": 200.674,
|
| 162 |
+
"eval_steps_per_second": 25.095,
|
| 163 |
+
"step": 7000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.12534469791927802,
|
| 167 |
+
"grad_norm": 16.960010528564453,
|
| 168 |
+
"learning_rate": 4.410550265441982e-05,
|
| 169 |
+
"loss": 3.8291,
|
| 170 |
+
"step": 7500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.13370101111389654,
|
| 174 |
+
"grad_norm": 15.129623413085938,
|
| 175 |
+
"learning_rate": 4.3684166175107444e-05,
|
| 176 |
+
"loss": 3.7697,
|
| 177 |
+
"step": 8000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.13370101111389654,
|
| 181 |
+
"eval_loss": 3.777130126953125,
|
| 182 |
+
"eval_runtime": 22.2765,
|
| 183 |
+
"eval_samples_per_second": 200.3,
|
| 184 |
+
"eval_steps_per_second": 25.049,
|
| 185 |
+
"step": 8000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1420573243085151,
|
| 189 |
+
"grad_norm": 18.825756072998047,
|
| 190 |
+
"learning_rate": 4.326535771467094e-05,
|
| 191 |
+
"loss": 3.7714,
|
| 192 |
+
"step": 8500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1504136375031336,
|
| 196 |
+
"grad_norm": 13.93099308013916,
|
| 197 |
+
"learning_rate": 4.284402123535856e-05,
|
| 198 |
+
"loss": 3.7719,
|
| 199 |
+
"step": 9000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1504136375031336,
|
| 203 |
+
"eval_loss": 3.747012138366699,
|
| 204 |
+
"eval_runtime": 22.2951,
|
| 205 |
+
"eval_samples_per_second": 200.134,
|
| 206 |
+
"eval_steps_per_second": 25.028,
|
| 207 |
+
"step": 9000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.15876995069775215,
|
| 211 |
+
"grad_norm": 12.885889053344727,
|
| 212 |
+
"learning_rate": 4.2422684756046185e-05,
|
| 213 |
+
"loss": 3.7127,
|
| 214 |
+
"step": 9500
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.16712626389237067,
|
| 218 |
+
"grad_norm": 15.362942695617676,
|
| 219 |
+
"learning_rate": 4.200134827673381e-05,
|
| 220 |
+
"loss": 3.7135,
|
| 221 |
+
"step": 10000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.16712626389237067,
|
| 225 |
+
"eval_loss": 3.7190206050872803,
|
| 226 |
+
"eval_runtime": 22.2344,
|
| 227 |
+
"eval_samples_per_second": 200.68,
|
| 228 |
+
"eval_steps_per_second": 25.096,
|
| 229 |
+
"step": 10000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.17548257708698922,
|
| 233 |
+
"grad_norm": 12.432964324951172,
|
| 234 |
+
"learning_rate": 4.158001179742142e-05,
|
| 235 |
+
"loss": 3.6715,
|
| 236 |
+
"step": 10500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.18383889028160774,
|
| 240 |
+
"grad_norm": 25.97146987915039,
|
| 241 |
+
"learning_rate": 4.115867531810904e-05,
|
| 242 |
+
"loss": 3.6599,
|
| 243 |
+
"step": 11000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.18383889028160774,
|
| 247 |
+
"eval_loss": 3.6818652153015137,
|
| 248 |
+
"eval_runtime": 22.2972,
|
| 249 |
+
"eval_samples_per_second": 200.115,
|
| 250 |
+
"eval_steps_per_second": 25.026,
|
| 251 |
+
"step": 11000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.1921952034762263,
|
| 255 |
+
"grad_norm": 21.378082275390625,
|
| 256 |
+
"learning_rate": 4.073733883879666e-05,
|
| 257 |
+
"loss": 3.6754,
|
| 258 |
+
"step": 11500
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.20055151667084484,
|
| 262 |
+
"grad_norm": 11.59192943572998,
|
| 263 |
+
"learning_rate": 4.031600235948428e-05,
|
| 264 |
+
"loss": 3.669,
|
| 265 |
+
"step": 12000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.20055151667084484,
|
| 269 |
+
"eval_loss": 3.661188840866089,
|
| 270 |
+
"eval_runtime": 22.2553,
|
| 271 |
+
"eval_samples_per_second": 200.492,
|
| 272 |
+
"eval_steps_per_second": 25.073,
|
| 273 |
+
"step": 12000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.20890782986546336,
|
| 277 |
+
"grad_norm": 17.75707244873047,
|
| 278 |
+
"learning_rate": 3.9894665880171905e-05,
|
| 279 |
+
"loss": 3.6794,
|
| 280 |
+
"step": 12500
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.2172641430600819,
|
| 284 |
+
"grad_norm": 10.825678825378418,
|
| 285 |
+
"learning_rate": 3.947332940085953e-05,
|
| 286 |
+
"loss": 3.6113,
|
| 287 |
+
"step": 13000
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.2172641430600819,
|
| 291 |
+
"eval_loss": 3.6508119106292725,
|
| 292 |
+
"eval_runtime": 22.2982,
|
| 293 |
+
"eval_samples_per_second": 200.106,
|
| 294 |
+
"eval_steps_per_second": 25.024,
|
| 295 |
+
"step": 13000
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.22562045625470042,
|
| 299 |
+
"grad_norm": 10.04261302947998,
|
| 300 |
+
"learning_rate": 3.905199292154715e-05,
|
| 301 |
+
"loss": 3.5963,
|
| 302 |
+
"step": 13500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.23397676944931897,
|
| 306 |
+
"grad_norm": 13.926618576049805,
|
| 307 |
+
"learning_rate": 3.863065644223477e-05,
|
| 308 |
+
"loss": 3.5997,
|
| 309 |
+
"step": 14000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.23397676944931897,
|
| 313 |
+
"eval_loss": 3.6223905086517334,
|
| 314 |
+
"eval_runtime": 22.2717,
|
| 315 |
+
"eval_samples_per_second": 200.344,
|
| 316 |
+
"eval_steps_per_second": 25.054,
|
| 317 |
+
"step": 14000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.2423330826439375,
|
| 321 |
+
"grad_norm": 11.00304889678955,
|
| 322 |
+
"learning_rate": 3.820931996292239e-05,
|
| 323 |
+
"loss": 3.5991,
|
| 324 |
+
"step": 14500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.25068939583855604,
|
| 328 |
+
"grad_norm": 16.099769592285156,
|
| 329 |
+
"learning_rate": 3.778798348361001e-05,
|
| 330 |
+
"loss": 3.6042,
|
| 331 |
+
"step": 15000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.25068939583855604,
|
| 335 |
+
"eval_loss": 3.5953731536865234,
|
| 336 |
+
"eval_runtime": 22.2814,
|
| 337 |
+
"eval_samples_per_second": 200.257,
|
| 338 |
+
"eval_steps_per_second": 25.043,
|
| 339 |
+
"step": 15000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2590457090331746,
|
| 343 |
+
"grad_norm": 12.459487915039062,
|
| 344 |
+
"learning_rate": 3.736664700429763e-05,
|
| 345 |
+
"loss": 3.5871,
|
| 346 |
+
"step": 15500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2674020222277931,
|
| 350 |
+
"grad_norm": 16.979909896850586,
|
| 351 |
+
"learning_rate": 3.6946153197943875e-05,
|
| 352 |
+
"loss": 3.5238,
|
| 353 |
+
"step": 16000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2674020222277931,
|
| 357 |
+
"eval_loss": 3.590113401412964,
|
| 358 |
+
"eval_runtime": 22.2293,
|
| 359 |
+
"eval_samples_per_second": 200.726,
|
| 360 |
+
"eval_steps_per_second": 25.102,
|
| 361 |
+
"step": 16000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.2757583354224116,
|
| 365 |
+
"grad_norm": 23.20758056640625,
|
| 366 |
+
"learning_rate": 3.65248167186315e-05,
|
| 367 |
+
"loss": 3.5646,
|
| 368 |
+
"step": 16500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2841146486170302,
|
| 372 |
+
"grad_norm": 18.35931396484375,
|
| 373 |
+
"learning_rate": 3.610348023931912e-05,
|
| 374 |
+
"loss": 3.5445,
|
| 375 |
+
"step": 17000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.2841146486170302,
|
| 379 |
+
"eval_loss": 3.563676595687866,
|
| 380 |
+
"eval_runtime": 22.2571,
|
| 381 |
+
"eval_samples_per_second": 200.475,
|
| 382 |
+
"eval_steps_per_second": 25.071,
|
| 383 |
+
"step": 17000
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.2924709618116487,
|
| 387 |
+
"grad_norm": 17.187950134277344,
|
| 388 |
+
"learning_rate": 3.568214376000674e-05,
|
| 389 |
+
"loss": 3.494,
|
| 390 |
+
"step": 17500
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.3008272750062672,
|
| 394 |
+
"grad_norm": 15.331987380981445,
|
| 395 |
+
"learning_rate": 3.5261649953652984e-05,
|
| 396 |
+
"loss": 3.4913,
|
| 397 |
+
"step": 18000
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.3008272750062672,
|
| 401 |
+
"eval_loss": 3.541306495666504,
|
| 402 |
+
"eval_runtime": 22.2598,
|
| 403 |
+
"eval_samples_per_second": 200.451,
|
| 404 |
+
"eval_steps_per_second": 25.068,
|
| 405 |
+
"step": 18000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.30918358820088576,
|
| 409 |
+
"grad_norm": 16.340852737426758,
|
| 410 |
+
"learning_rate": 3.484031347434061e-05,
|
| 411 |
+
"loss": 3.4969,
|
| 412 |
+
"step": 18500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.3175399013955043,
|
| 416 |
+
"grad_norm": 12.265207290649414,
|
| 417 |
+
"learning_rate": 3.441897699502823e-05,
|
| 418 |
+
"loss": 3.4934,
|
| 419 |
+
"step": 19000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.3175399013955043,
|
| 423 |
+
"eval_loss": 3.520357847213745,
|
| 424 |
+
"eval_runtime": 22.2447,
|
| 425 |
+
"eval_samples_per_second": 200.587,
|
| 426 |
+
"eval_steps_per_second": 25.085,
|
| 427 |
+
"step": 19000
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.32589621459012286,
|
| 431 |
+
"grad_norm": 15.456232070922852,
|
| 432 |
+
"learning_rate": 3.399764051571585e-05,
|
| 433 |
+
"loss": 3.5013,
|
| 434 |
+
"step": 19500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.33425252778474135,
|
| 438 |
+
"grad_norm": 15.721699714660645,
|
| 439 |
+
"learning_rate": 3.3576304036403474e-05,
|
| 440 |
+
"loss": 3.4627,
|
| 441 |
+
"step": 20000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.33425252778474135,
|
| 445 |
+
"eval_loss": 3.5179378986358643,
|
| 446 |
+
"eval_runtime": 22.2594,
|
| 447 |
+
"eval_samples_per_second": 200.455,
|
| 448 |
+
"eval_steps_per_second": 25.068,
|
| 449 |
+
"step": 20000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.3426088409793599,
|
| 453 |
+
"grad_norm": 12.118553161621094,
|
| 454 |
+
"learning_rate": 3.3154967557091096e-05,
|
| 455 |
+
"loss": 3.5006,
|
| 456 |
+
"step": 20500
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.35096515417397844,
|
| 460 |
+
"grad_norm": 8.990864753723145,
|
| 461 |
+
"learning_rate": 3.273447375073734e-05,
|
| 462 |
+
"loss": 3.4367,
|
| 463 |
+
"step": 21000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.35096515417397844,
|
| 467 |
+
"eval_loss": 3.5118658542633057,
|
| 468 |
+
"eval_runtime": 22.2415,
|
| 469 |
+
"eval_samples_per_second": 200.616,
|
| 470 |
+
"eval_steps_per_second": 25.088,
|
| 471 |
+
"step": 21000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.359321467368597,
|
| 475 |
+
"grad_norm": 9.9972562789917,
|
| 476 |
+
"learning_rate": 3.231313727142496e-05,
|
| 477 |
+
"loss": 3.4498,
|
| 478 |
+
"step": 21500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.3676777805632155,
|
| 482 |
+
"grad_norm": 10.996673583984375,
|
| 483 |
+
"learning_rate": 3.189180079211258e-05,
|
| 484 |
+
"loss": 3.4643,
|
| 485 |
+
"step": 22000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.3676777805632155,
|
| 489 |
+
"eval_loss": 3.483738899230957,
|
| 490 |
+
"eval_runtime": 22.2582,
|
| 491 |
+
"eval_samples_per_second": 200.465,
|
| 492 |
+
"eval_steps_per_second": 25.069,
|
| 493 |
+
"step": 22000
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.37603409375783403,
|
| 497 |
+
"grad_norm": 14.55636978149414,
|
| 498 |
+
"learning_rate": 3.14704643128002e-05,
|
| 499 |
+
"loss": 3.5215,
|
| 500 |
+
"step": 22500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3843904069524526,
|
| 504 |
+
"grad_norm": 13.585105895996094,
|
| 505 |
+
"learning_rate": 3.104912783348782e-05,
|
| 506 |
+
"loss": 3.419,
|
| 507 |
+
"step": 23000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.3843904069524526,
|
| 511 |
+
"eval_loss": 3.47660231590271,
|
| 512 |
+
"eval_runtime": 22.226,
|
| 513 |
+
"eval_samples_per_second": 200.756,
|
| 514 |
+
"eval_steps_per_second": 25.106,
|
| 515 |
+
"step": 23000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.3927467201470711,
|
| 519 |
+
"grad_norm": 11.853238105773926,
|
| 520 |
+
"learning_rate": 3.062779135417544e-05,
|
| 521 |
+
"loss": 3.4438,
|
| 522 |
+
"step": 23500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.4011030333416897,
|
| 526 |
+
"grad_norm": 13.06174373626709,
|
| 527 |
+
"learning_rate": 3.020729754782169e-05,
|
| 528 |
+
"loss": 3.4029,
|
| 529 |
+
"step": 24000
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.4011030333416897,
|
| 533 |
+
"eval_loss": 3.4587268829345703,
|
| 534 |
+
"eval_runtime": 22.2726,
|
| 535 |
+
"eval_samples_per_second": 200.336,
|
| 536 |
+
"eval_steps_per_second": 25.053,
|
| 537 |
+
"step": 24000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.40945934653630817,
|
| 541 |
+
"grad_norm": 16.874757766723633,
|
| 542 |
+
"learning_rate": 2.9786803741467938e-05,
|
| 543 |
+
"loss": 3.3971,
|
| 544 |
+
"step": 24500
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.4178156597309267,
|
| 548 |
+
"grad_norm": 11.108474731445312,
|
| 549 |
+
"learning_rate": 2.936546726215556e-05,
|
| 550 |
+
"loss": 3.3574,
|
| 551 |
+
"step": 25000
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.4178156597309267,
|
| 555 |
+
"eval_loss": 3.446179151535034,
|
| 556 |
+
"eval_runtime": 22.3522,
|
| 557 |
+
"eval_samples_per_second": 199.622,
|
| 558 |
+
"eval_steps_per_second": 24.964,
|
| 559 |
+
"step": 25000
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.42617197292554526,
|
| 563 |
+
"grad_norm": 12.936110496520996,
|
| 564 |
+
"learning_rate": 2.8944130782843183e-05,
|
| 565 |
+
"loss": 3.3829,
|
| 566 |
+
"step": 25500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.4345282861201638,
|
| 570 |
+
"grad_norm": 12.90854549407959,
|
| 571 |
+
"learning_rate": 2.8522794303530802e-05,
|
| 572 |
+
"loss": 3.4156,
|
| 573 |
+
"step": 26000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.4345282861201638,
|
| 577 |
+
"eval_loss": 3.44026517868042,
|
| 578 |
+
"eval_runtime": 22.2661,
|
| 579 |
+
"eval_samples_per_second": 200.394,
|
| 580 |
+
"eval_steps_per_second": 25.061,
|
| 581 |
+
"step": 26000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.4428845993147823,
|
| 585 |
+
"grad_norm": 10.326555252075195,
|
| 586 |
+
"learning_rate": 2.8101457824218424e-05,
|
| 587 |
+
"loss": 3.3607,
|
| 588 |
+
"step": 26500
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.45124091250940085,
|
| 592 |
+
"grad_norm": 12.372066497802734,
|
| 593 |
+
"learning_rate": 2.7681806690823293e-05,
|
| 594 |
+
"loss": 3.3836,
|
| 595 |
+
"step": 27000
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.45124091250940085,
|
| 599 |
+
"eval_loss": 3.4253649711608887,
|
| 600 |
+
"eval_runtime": 22.2507,
|
| 601 |
+
"eval_samples_per_second": 200.533,
|
| 602 |
+
"eval_steps_per_second": 25.078,
|
| 603 |
+
"step": 27000
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.4595972257040194,
|
| 607 |
+
"grad_norm": 9.778299331665039,
|
| 608 |
+
"learning_rate": 2.7260470211510912e-05,
|
| 609 |
+
"loss": 3.3671,
|
| 610 |
+
"step": 27500
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.46795353889863794,
|
| 614 |
+
"grad_norm": 20.047178268432617,
|
| 615 |
+
"learning_rate": 2.6839133732198535e-05,
|
| 616 |
+
"loss": 3.3395,
|
| 617 |
+
"step": 28000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.46795353889863794,
|
| 621 |
+
"eval_loss": 3.41679048538208,
|
| 622 |
+
"eval_runtime": 22.2707,
|
| 623 |
+
"eval_samples_per_second": 200.353,
|
| 624 |
+
"eval_steps_per_second": 25.055,
|
| 625 |
+
"step": 28000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.47630985209325644,
|
| 629 |
+
"grad_norm": 9.312335968017578,
|
| 630 |
+
"learning_rate": 2.6417797252886157e-05,
|
| 631 |
+
"loss": 3.3616,
|
| 632 |
+
"step": 28500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.484666165287875,
|
| 636 |
+
"grad_norm": 10.994682312011719,
|
| 637 |
+
"learning_rate": 2.5996460773573776e-05,
|
| 638 |
+
"loss": 3.3719,
|
| 639 |
+
"step": 29000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.484666165287875,
|
| 643 |
+
"eval_loss": 3.4018924236297607,
|
| 644 |
+
"eval_runtime": 22.2565,
|
| 645 |
+
"eval_samples_per_second": 200.481,
|
| 646 |
+
"eval_steps_per_second": 25.071,
|
| 647 |
+
"step": 29000
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.49302247848249353,
|
| 651 |
+
"grad_norm": 13.464505195617676,
|
| 652 |
+
"learning_rate": 2.5575124294261398e-05,
|
| 653 |
+
"loss": 3.3312,
|
| 654 |
+
"step": 29500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5013787916771121,
|
| 658 |
+
"grad_norm": 12.18619441986084,
|
| 659 |
+
"learning_rate": 2.515378781494902e-05,
|
| 660 |
+
"loss": 3.386,
|
| 661 |
+
"step": 30000
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5013787916771121,
|
| 665 |
+
"eval_loss": 3.3899354934692383,
|
| 666 |
+
"eval_runtime": 22.2658,
|
| 667 |
+
"eval_samples_per_second": 200.397,
|
| 668 |
+
"eval_steps_per_second": 25.061,
|
| 669 |
+
"step": 30000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.5097351048717306,
|
| 673 |
+
"grad_norm": 14.552848815917969,
|
| 674 |
+
"learning_rate": 2.4732451335636643e-05,
|
| 675 |
+
"loss": 3.3377,
|
| 676 |
+
"step": 30500
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.5180914180663492,
|
| 680 |
+
"grad_norm": 15.032088279724121,
|
| 681 |
+
"learning_rate": 2.4311114856324262e-05,
|
| 682 |
+
"loss": 3.3131,
|
| 683 |
+
"step": 31000
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.5180914180663492,
|
| 687 |
+
"eval_loss": 3.378127336502075,
|
| 688 |
+
"eval_runtime": 22.2423,
|
| 689 |
+
"eval_samples_per_second": 200.609,
|
| 690 |
+
"eval_steps_per_second": 25.087,
|
| 691 |
+
"step": 31000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.5264477312609677,
|
| 695 |
+
"grad_norm": 14.666757583618164,
|
| 696 |
+
"learning_rate": 2.388977837701188e-05,
|
| 697 |
+
"loss": 3.3457,
|
| 698 |
+
"step": 31500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.5348040444555862,
|
| 702 |
+
"grad_norm": 11.800482749938965,
|
| 703 |
+
"learning_rate": 2.3468441897699503e-05,
|
| 704 |
+
"loss": 3.3192,
|
| 705 |
+
"step": 32000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.5348040444555862,
|
| 709 |
+
"eval_loss": 3.3670458793640137,
|
| 710 |
+
"eval_runtime": 22.256,
|
| 711 |
+
"eval_samples_per_second": 200.485,
|
| 712 |
+
"eval_steps_per_second": 25.072,
|
| 713 |
+
"step": 32000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.5431603576502048,
|
| 717 |
+
"grad_norm": 10.835103034973145,
|
| 718 |
+
"learning_rate": 2.3047105418387125e-05,
|
| 719 |
+
"loss": 3.3235,
|
| 720 |
+
"step": 32500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.5515166708448233,
|
| 724 |
+
"grad_norm": 12.06092357635498,
|
| 725 |
+
"learning_rate": 2.2625768939074744e-05,
|
| 726 |
+
"loss": 3.2969,
|
| 727 |
+
"step": 33000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.5515166708448233,
|
| 731 |
+
"eval_loss": 3.356658935546875,
|
| 732 |
+
"eval_runtime": 22.2404,
|
| 733 |
+
"eval_samples_per_second": 200.626,
|
| 734 |
+
"eval_steps_per_second": 25.089,
|
| 735 |
+
"step": 33000
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.5598729840394417,
|
| 739 |
+
"grad_norm": 15.398877143859863,
|
| 740 |
+
"learning_rate": 2.2204432459762367e-05,
|
| 741 |
+
"loss": 3.3181,
|
| 742 |
+
"step": 33500
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.5682292972340603,
|
| 746 |
+
"grad_norm": 10.425477027893066,
|
| 747 |
+
"learning_rate": 2.178309598044999e-05,
|
| 748 |
+
"loss": 3.3202,
|
| 749 |
+
"step": 34000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.5682292972340603,
|
| 753 |
+
"eval_loss": 3.34324312210083,
|
| 754 |
+
"eval_runtime": 22.2237,
|
| 755 |
+
"eval_samples_per_second": 200.777,
|
| 756 |
+
"eval_steps_per_second": 25.108,
|
| 757 |
+
"step": 34000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.5765856104286788,
|
| 761 |
+
"grad_norm": 13.118115425109863,
|
| 762 |
+
"learning_rate": 2.136175950113761e-05,
|
| 763 |
+
"loss": 3.3028,
|
| 764 |
+
"step": 34500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.5849419236232974,
|
| 768 |
+
"grad_norm": 8.235157012939453,
|
| 769 |
+
"learning_rate": 2.0941265694783854e-05,
|
| 770 |
+
"loss": 3.2403,
|
| 771 |
+
"step": 35000
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.5849419236232974,
|
| 775 |
+
"eval_loss": 3.3430681228637695,
|
| 776 |
+
"eval_runtime": 22.2974,
|
| 777 |
+
"eval_samples_per_second": 200.113,
|
| 778 |
+
"eval_steps_per_second": 25.025,
|
| 779 |
+
"step": 35000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5932982368179159,
|
| 783 |
+
"grad_norm": 15.389208793640137,
|
| 784 |
+
"learning_rate": 2.0519929215471476e-05,
|
| 785 |
+
"loss": 3.3105,
|
| 786 |
+
"step": 35500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.6016545500125344,
|
| 790 |
+
"grad_norm": 12.708732604980469,
|
| 791 |
+
"learning_rate": 2.0098592736159098e-05,
|
| 792 |
+
"loss": 3.2775,
|
| 793 |
+
"step": 36000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6016545500125344,
|
| 797 |
+
"eval_loss": 3.3276991844177246,
|
| 798 |
+
"eval_runtime": 22.2643,
|
| 799 |
+
"eval_samples_per_second": 200.411,
|
| 800 |
+
"eval_steps_per_second": 25.063,
|
| 801 |
+
"step": 36000
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.610010863207153,
|
| 805 |
+
"grad_norm": 13.642451286315918,
|
| 806 |
+
"learning_rate": 1.9677256256846717e-05,
|
| 807 |
+
"loss": 3.2902,
|
| 808 |
+
"step": 36500
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6183671764017715,
|
| 812 |
+
"grad_norm": 12.606600761413574,
|
| 813 |
+
"learning_rate": 1.9256762450492966e-05,
|
| 814 |
+
"loss": 3.271,
|
| 815 |
+
"step": 37000
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.6183671764017715,
|
| 819 |
+
"eval_loss": 3.3122496604919434,
|
| 820 |
+
"eval_runtime": 22.2804,
|
| 821 |
+
"eval_samples_per_second": 200.266,
|
| 822 |
+
"eval_steps_per_second": 25.044,
|
| 823 |
+
"step": 37000
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.6267234895963901,
|
| 827 |
+
"grad_norm": 11.484159469604492,
|
| 828 |
+
"learning_rate": 1.8835425971180585e-05,
|
| 829 |
+
"loss": 3.2833,
|
| 830 |
+
"step": 37500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.6350798027910086,
|
| 834 |
+
"grad_norm": 12.317131996154785,
|
| 835 |
+
"learning_rate": 1.8414089491868204e-05,
|
| 836 |
+
"loss": 3.2848,
|
| 837 |
+
"step": 38000
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.6350798027910086,
|
| 841 |
+
"eval_loss": 3.3035213947296143,
|
| 842 |
+
"eval_runtime": 22.2937,
|
| 843 |
+
"eval_samples_per_second": 200.147,
|
| 844 |
+
"eval_steps_per_second": 25.03,
|
| 845 |
+
"step": 38000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.6434361159856271,
|
| 849 |
+
"grad_norm": 11.45077896118164,
|
| 850 |
+
"learning_rate": 1.7992753012555827e-05,
|
| 851 |
+
"loss": 3.202,
|
| 852 |
+
"step": 38500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.6517924291802457,
|
| 856 |
+
"grad_norm": 12.859657287597656,
|
| 857 |
+
"learning_rate": 1.7572259206202076e-05,
|
| 858 |
+
"loss": 3.2376,
|
| 859 |
+
"step": 39000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.6517924291802457,
|
| 863 |
+
"eval_loss": 3.2956559658050537,
|
| 864 |
+
"eval_runtime": 22.3804,
|
| 865 |
+
"eval_samples_per_second": 199.371,
|
| 866 |
+
"eval_steps_per_second": 24.933,
|
| 867 |
+
"step": 39000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.6601487423748642,
|
| 871 |
+
"grad_norm": 14.472012519836426,
|
| 872 |
+
"learning_rate": 1.7150922726889695e-05,
|
| 873 |
+
"loss": 3.1924,
|
| 874 |
+
"step": 39500
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.6685050555694827,
|
| 878 |
+
"grad_norm": 13.051079750061035,
|
| 879 |
+
"learning_rate": 1.673042892053594e-05,
|
| 880 |
+
"loss": 3.2598,
|
| 881 |
+
"step": 40000
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.6685050555694827,
|
| 885 |
+
"eval_loss": 3.2878499031066895,
|
| 886 |
+
"eval_runtime": 22.2464,
|
| 887 |
+
"eval_samples_per_second": 200.572,
|
| 888 |
+
"eval_steps_per_second": 25.083,
|
| 889 |
+
"step": 40000
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.6768613687641013,
|
| 893 |
+
"grad_norm": 15.44560718536377,
|
| 894 |
+
"learning_rate": 1.6309092441223563e-05,
|
| 895 |
+
"loss": 3.1978,
|
| 896 |
+
"step": 40500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.6852176819587198,
|
| 900 |
+
"grad_norm": 16.988996505737305,
|
| 901 |
+
"learning_rate": 1.588775596191118e-05,
|
| 902 |
+
"loss": 3.2247,
|
| 903 |
+
"step": 41000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.6852176819587198,
|
| 907 |
+
"eval_loss": 3.279550313949585,
|
| 908 |
+
"eval_runtime": 22.2386,
|
| 909 |
+
"eval_samples_per_second": 200.642,
|
| 910 |
+
"eval_steps_per_second": 25.091,
|
| 911 |
+
"step": 41000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.6935739951533384,
|
| 915 |
+
"grad_norm": 8.293917655944824,
|
| 916 |
+
"learning_rate": 1.5466419482598804e-05,
|
| 917 |
+
"loss": 3.1682,
|
| 918 |
+
"step": 41500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.7019303083479569,
|
| 922 |
+
"grad_norm": 10.755880355834961,
|
| 923 |
+
"learning_rate": 1.5045925676245051e-05,
|
| 924 |
+
"loss": 3.1849,
|
| 925 |
+
"step": 42000
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.7019303083479569,
|
| 929 |
+
"eval_loss": 3.2791192531585693,
|
| 930 |
+
"eval_runtime": 22.2554,
|
| 931 |
+
"eval_samples_per_second": 200.491,
|
| 932 |
+
"eval_steps_per_second": 25.073,
|
| 933 |
+
"step": 42000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.7102866215425754,
|
| 937 |
+
"grad_norm": 17.822643280029297,
|
| 938 |
+
"learning_rate": 1.462458919693267e-05,
|
| 939 |
+
"loss": 3.1714,
|
| 940 |
+
"step": 42500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.718642934737194,
|
| 944 |
+
"grad_norm": 18.230485916137695,
|
| 945 |
+
"learning_rate": 1.4203252717620291e-05,
|
| 946 |
+
"loss": 3.2112,
|
| 947 |
+
"step": 43000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.718642934737194,
|
| 951 |
+
"eval_loss": 3.260193109512329,
|
| 952 |
+
"eval_runtime": 22.2518,
|
| 953 |
+
"eval_samples_per_second": 200.523,
|
| 954 |
+
"eval_steps_per_second": 25.077,
|
| 955 |
+
"step": 43000
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.7269992479318125,
|
| 959 |
+
"grad_norm": 13.363430976867676,
|
| 960 |
+
"learning_rate": 1.3781916238307913e-05,
|
| 961 |
+
"loss": 3.1655,
|
| 962 |
+
"step": 43500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.735355561126431,
|
| 966 |
+
"grad_norm": 11.570181846618652,
|
| 967 |
+
"learning_rate": 1.3360579758995534e-05,
|
| 968 |
+
"loss": 3.174,
|
| 969 |
+
"step": 44000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.735355561126431,
|
| 973 |
+
"eval_loss": 3.2490386962890625,
|
| 974 |
+
"eval_runtime": 22.2683,
|
| 975 |
+
"eval_samples_per_second": 200.374,
|
| 976 |
+
"eval_steps_per_second": 25.058,
|
| 977 |
+
"step": 44000
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.7437118743210496,
|
| 981 |
+
"grad_norm": 19.80602264404297,
|
| 982 |
+
"learning_rate": 1.2939243279683155e-05,
|
| 983 |
+
"loss": 3.1987,
|
| 984 |
+
"step": 44500
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.7520681875156681,
|
| 988 |
+
"grad_norm": 10.821731567382812,
|
| 989 |
+
"learning_rate": 1.2518749473329402e-05,
|
| 990 |
+
"loss": 3.1799,
|
| 991 |
+
"step": 45000
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.7520681875156681,
|
| 995 |
+
"eval_loss": 3.240847587585449,
|
| 996 |
+
"eval_runtime": 22.2794,
|
| 997 |
+
"eval_samples_per_second": 200.275,
|
| 998 |
+
"eval_steps_per_second": 25.046,
|
| 999 |
+
"step": 45000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.7604245007102867,
|
| 1003 |
+
"grad_norm": 16.301612854003906,
|
| 1004 |
+
"learning_rate": 1.2097412994017023e-05,
|
| 1005 |
+
"loss": 3.2029,
|
| 1006 |
+
"step": 45500
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.7687808139049052,
|
| 1010 |
+
"grad_norm": 14.699359893798828,
|
| 1011 |
+
"learning_rate": 1.1676076514704643e-05,
|
| 1012 |
+
"loss": 3.1752,
|
| 1013 |
+
"step": 46000
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.7687808139049052,
|
| 1017 |
+
"eval_loss": 3.233914852142334,
|
| 1018 |
+
"eval_runtime": 22.269,
|
| 1019 |
+
"eval_samples_per_second": 200.369,
|
| 1020 |
+
"eval_steps_per_second": 25.057,
|
| 1021 |
+
"step": 46000
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 0.7771371270995237,
|
| 1025 |
+
"grad_norm": 15.696563720703125,
|
| 1026 |
+
"learning_rate": 1.1254740035392266e-05,
|
| 1027 |
+
"loss": 3.132,
|
| 1028 |
+
"step": 46500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.7854934402941423,
|
| 1032 |
+
"grad_norm": 13.062487602233887,
|
| 1033 |
+
"learning_rate": 1.0833403556079886e-05,
|
| 1034 |
+
"loss": 3.131,
|
| 1035 |
+
"step": 47000
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.7854934402941423,
|
| 1039 |
+
"eval_loss": 3.2280752658843994,
|
| 1040 |
+
"eval_runtime": 22.2955,
|
| 1041 |
+
"eval_samples_per_second": 200.13,
|
| 1042 |
+
"eval_steps_per_second": 25.027,
|
| 1043 |
+
"step": 47000
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.7938497534887607,
|
| 1047 |
+
"grad_norm": 18.67305564880371,
|
| 1048 |
+
"learning_rate": 1.0412909749726132e-05,
|
| 1049 |
+
"loss": 3.1571,
|
| 1050 |
+
"step": 47500
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.8022060666833793,
|
| 1054 |
+
"grad_norm": 10.377827644348145,
|
| 1055 |
+
"learning_rate": 9.992415943372378e-06,
|
| 1056 |
+
"loss": 3.181,
|
| 1057 |
+
"step": 48000
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.8022060666833793,
|
| 1061 |
+
"eval_loss": 3.2206084728240967,
|
| 1062 |
+
"eval_runtime": 22.2943,
|
| 1063 |
+
"eval_samples_per_second": 200.141,
|
| 1064 |
+
"eval_steps_per_second": 25.029,
|
| 1065 |
+
"step": 48000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.8105623798779978,
|
| 1069 |
+
"grad_norm": 12.836233139038086,
|
| 1070 |
+
"learning_rate": 9.571922137018624e-06,
|
| 1071 |
+
"loss": 3.139,
|
| 1072 |
+
"step": 48500
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.8189186930726163,
|
| 1076 |
+
"grad_norm": 11.736408233642578,
|
| 1077 |
+
"learning_rate": 9.150585657706244e-06,
|
| 1078 |
+
"loss": 3.0932,
|
| 1079 |
+
"step": 49000
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.8189186930726163,
|
| 1083 |
+
"eval_loss": 3.2135069370269775,
|
| 1084 |
+
"eval_runtime": 22.2506,
|
| 1085 |
+
"eval_samples_per_second": 200.534,
|
| 1086 |
+
"eval_steps_per_second": 25.078,
|
| 1087 |
+
"step": 49000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.8272750062672349,
|
| 1091 |
+
"grad_norm": 16.016298294067383,
|
| 1092 |
+
"learning_rate": 8.729249178393865e-06,
|
| 1093 |
+
"loss": 3.1634,
|
| 1094 |
+
"step": 49500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.8356313194618534,
|
| 1098 |
+
"grad_norm": 10.488819122314453,
|
| 1099 |
+
"learning_rate": 8.307912699081487e-06,
|
| 1100 |
+
"loss": 3.1376,
|
| 1101 |
+
"step": 50000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.8356313194618534,
|
| 1105 |
+
"eval_loss": 3.2051162719726562,
|
| 1106 |
+
"eval_runtime": 22.294,
|
| 1107 |
+
"eval_samples_per_second": 200.144,
|
| 1108 |
+
"eval_steps_per_second": 25.029,
|
| 1109 |
+
"step": 50000
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.8439876326564719,
|
| 1113 |
+
"grad_norm": 16.168071746826172,
|
| 1114 |
+
"learning_rate": 7.886576219769108e-06,
|
| 1115 |
+
"loss": 3.1121,
|
| 1116 |
+
"step": 50500
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.8523439458510905,
|
| 1120 |
+
"grad_norm": 19.903099060058594,
|
| 1121 |
+
"learning_rate": 7.465239740456729e-06,
|
| 1122 |
+
"loss": 3.1084,
|
| 1123 |
+
"step": 51000
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.8523439458510905,
|
| 1127 |
+
"eval_loss": 3.198310375213623,
|
| 1128 |
+
"eval_runtime": 22.3049,
|
| 1129 |
+
"eval_samples_per_second": 200.046,
|
| 1130 |
+
"eval_steps_per_second": 25.017,
|
| 1131 |
+
"step": 51000
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.860700259045709,
|
| 1135 |
+
"grad_norm": 12.082676887512207,
|
| 1136 |
+
"learning_rate": 7.043903261144351e-06,
|
| 1137 |
+
"loss": 3.0957,
|
| 1138 |
+
"step": 51500
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.8690565722403276,
|
| 1142 |
+
"grad_norm": 11.764552116394043,
|
| 1143 |
+
"learning_rate": 6.622566781831971e-06,
|
| 1144 |
+
"loss": 3.099,
|
| 1145 |
+
"step": 52000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.8690565722403276,
|
| 1149 |
+
"eval_loss": 3.193253993988037,
|
| 1150 |
+
"eval_runtime": 22.2469,
|
| 1151 |
+
"eval_samples_per_second": 200.567,
|
| 1152 |
+
"eval_steps_per_second": 25.082,
|
| 1153 |
+
"step": 52000
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.8774128854349461,
|
| 1157 |
+
"grad_norm": 12.482972145080566,
|
| 1158 |
+
"learning_rate": 6.201230302519592e-06,
|
| 1159 |
+
"loss": 3.0779,
|
| 1160 |
+
"step": 52500
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.8857691986295646,
|
| 1164 |
+
"grad_norm": 14.11436939239502,
|
| 1165 |
+
"learning_rate": 5.7798938232072135e-06,
|
| 1166 |
+
"loss": 3.1278,
|
| 1167 |
+
"step": 53000
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.8857691986295646,
|
| 1171 |
+
"eval_loss": 3.1867904663085938,
|
| 1172 |
+
"eval_runtime": 22.2664,
|
| 1173 |
+
"eval_samples_per_second": 200.392,
|
| 1174 |
+
"eval_steps_per_second": 25.06,
|
| 1175 |
+
"step": 53000
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.8941255118241832,
|
| 1179 |
+
"grad_norm": 19.69700813293457,
|
| 1180 |
+
"learning_rate": 5.358557343894835e-06,
|
| 1181 |
+
"loss": 3.0968,
|
| 1182 |
+
"step": 53500
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.9024818250188017,
|
| 1186 |
+
"grad_norm": 14.537339210510254,
|
| 1187 |
+
"learning_rate": 4.937220864582456e-06,
|
| 1188 |
+
"loss": 3.1436,
|
| 1189 |
+
"step": 54000
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.9024818250188017,
|
| 1193 |
+
"eval_loss": 3.180774688720703,
|
| 1194 |
+
"eval_runtime": 22.293,
|
| 1195 |
+
"eval_samples_per_second": 200.152,
|
| 1196 |
+
"eval_steps_per_second": 25.03,
|
| 1197 |
+
"step": 54000
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.9108381382134202,
|
| 1201 |
+
"grad_norm": 16.117996215820312,
|
| 1202 |
+
"learning_rate": 4.515884385270077e-06,
|
| 1203 |
+
"loss": 3.1288,
|
| 1204 |
+
"step": 54500
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 0.9191944514080388,
|
| 1208 |
+
"grad_norm": 12.458276748657227,
|
| 1209 |
+
"learning_rate": 4.094547905957698e-06,
|
| 1210 |
+
"loss": 3.0763,
|
| 1211 |
+
"step": 55000
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.9191944514080388,
|
| 1215 |
+
"eval_loss": 3.175370216369629,
|
| 1216 |
+
"eval_runtime": 22.253,
|
| 1217 |
+
"eval_samples_per_second": 200.513,
|
| 1218 |
+
"eval_steps_per_second": 25.075,
|
| 1219 |
+
"step": 55000
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 0.9275507646026573,
|
| 1223 |
+
"grad_norm": 14.115385055541992,
|
| 1224 |
+
"learning_rate": 3.6732114266453192e-06,
|
| 1225 |
+
"loss": 3.0642,
|
| 1226 |
+
"step": 55500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.9359070777972759,
|
| 1230 |
+
"grad_norm": 19.65464210510254,
|
| 1231 |
+
"learning_rate": 3.2518749473329403e-06,
|
| 1232 |
+
"loss": 3.1248,
|
| 1233 |
+
"step": 56000
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.9359070777972759,
|
| 1237 |
+
"eval_loss": 3.1690962314605713,
|
| 1238 |
+
"eval_runtime": 22.3258,
|
| 1239 |
+
"eval_samples_per_second": 199.858,
|
| 1240 |
+
"eval_steps_per_second": 24.993,
|
| 1241 |
+
"step": 56000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.9442633909918944,
|
| 1245 |
+
"grad_norm": 11.953753471374512,
|
| 1246 |
+
"learning_rate": 2.831381140979186e-06,
|
| 1247 |
+
"loss": 3.1361,
|
| 1248 |
+
"step": 56500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.9526197041865129,
|
| 1252 |
+
"grad_norm": 10.821110725402832,
|
| 1253 |
+
"learning_rate": 2.4108873346254323e-06,
|
| 1254 |
+
"loss": 3.0418,
|
| 1255 |
+
"step": 57000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.9526197041865129,
|
| 1259 |
+
"eval_loss": 3.164776563644409,
|
| 1260 |
+
"eval_runtime": 22.2464,
|
| 1261 |
+
"eval_samples_per_second": 200.572,
|
| 1262 |
+
"eval_steps_per_second": 25.083,
|
| 1263 |
+
"step": 57000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.9609760173811315,
|
| 1267 |
+
"grad_norm": 11.476717948913574,
|
| 1268 |
+
"learning_rate": 1.9895508553130533e-06,
|
| 1269 |
+
"loss": 3.0504,
|
| 1270 |
+
"step": 57500
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.96933233057575,
|
| 1274 |
+
"grad_norm": 10.973363876342773,
|
| 1275 |
+
"learning_rate": 1.5682143760006742e-06,
|
| 1276 |
+
"loss": 3.0755,
|
| 1277 |
+
"step": 58000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.96933233057575,
|
| 1281 |
+
"eval_loss": 3.1603705883026123,
|
| 1282 |
+
"eval_runtime": 22.272,
|
| 1283 |
+
"eval_samples_per_second": 200.341,
|
| 1284 |
+
"eval_steps_per_second": 25.054,
|
| 1285 |
+
"step": 58000
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.9776886437703685,
|
| 1289 |
+
"grad_norm": 10.836787223815918,
|
| 1290 |
+
"learning_rate": 1.1468778966882954e-06,
|
| 1291 |
+
"loss": 3.1001,
|
| 1292 |
+
"step": 58500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.9860449569649871,
|
| 1296 |
+
"grad_norm": 13.901703834533691,
|
| 1297 |
+
"learning_rate": 7.255414173759165e-07,
|
| 1298 |
+
"loss": 3.0633,
|
| 1299 |
+
"step": 59000
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.9860449569649871,
|
| 1303 |
+
"eval_loss": 3.1587648391723633,
|
| 1304 |
+
"eval_runtime": 22.3994,
|
| 1305 |
+
"eval_samples_per_second": 199.202,
|
| 1306 |
+
"eval_steps_per_second": 24.911,
|
| 1307 |
+
"step": 59000
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.9944012701596056,
|
| 1311 |
+
"grad_norm": 19.022567749023438,
|
| 1312 |
+
"learning_rate": 3.0420493806353753e-07,
|
| 1313 |
+
"loss": 3.0751,
|
| 1314 |
+
"step": 59500
|
| 1315 |
+
}
|
| 1316 |
+
],
|
| 1317 |
+
"logging_steps": 500,
|
| 1318 |
+
"max_steps": 59835,
|
| 1319 |
+
"num_input_tokens_seen": 0,
|
| 1320 |
+
"num_train_epochs": 1,
|
| 1321 |
+
"save_steps": 2000,
|
| 1322 |
+
"stateful_callbacks": {
|
| 1323 |
+
"TrainerControl": {
|
| 1324 |
+
"args": {
|
| 1325 |
+
"should_epoch_stop": false,
|
| 1326 |
+
"should_evaluate": false,
|
| 1327 |
+
"should_log": false,
|
| 1328 |
+
"should_save": true,
|
| 1329 |
+
"should_training_stop": true
|
| 1330 |
+
},
|
| 1331 |
+
"attributes": {}
|
| 1332 |
+
}
|
| 1333 |
+
},
|
| 1334 |
+
"total_flos": 2.943723113325527e+17,
|
| 1335 |
+
"train_batch_size": 8,
|
| 1336 |
+
"trial_name": null,
|
| 1337 |
+
"trial_params": null
|
| 1338 |
+
}
|
full-finetuning/LLaMmlein_120M/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3489adac5f0635d0dfab0d3adcddd340b685c77de41ab94351691f318fbda79
|
| 3 |
+
size 5240
|
full-finetuning/german-gpt2/config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "dbmdz/german-gpt2",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.0,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"embd_pdrop": 0.0,
|
| 10 |
+
"eos_token_id": 50256,
|
| 11 |
+
"gradient_checkpointing": false,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"layer_norm_epsilon": 1e-05,
|
| 14 |
+
"model_type": "gpt2",
|
| 15 |
+
"n_ctx": 1024,
|
| 16 |
+
"n_embd": 768,
|
| 17 |
+
"n_head": 12,
|
| 18 |
+
"n_inner": null,
|
| 19 |
+
"n_layer": 12,
|
| 20 |
+
"n_positions": 1024,
|
| 21 |
+
"reorder_and_upcast_attn": false,
|
| 22 |
+
"resid_pdrop": 0.0,
|
| 23 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 24 |
+
"scale_attn_weights": true,
|
| 25 |
+
"summary_activation": null,
|
| 26 |
+
"summary_first_dropout": 0.1,
|
| 27 |
+
"summary_proj_to_labels": true,
|
| 28 |
+
"summary_type": "cls_index",
|
| 29 |
+
"summary_use_proj": true,
|
| 30 |
+
"task_specific_params": {
|
| 31 |
+
"text-generation": {
|
| 32 |
+
"do_sample": true,
|
| 33 |
+
"max_length": 50
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"torch_dtype": "float32",
|
| 37 |
+
"transformers_version": "4.46.0",
|
| 38 |
+
"use_cache": true,
|
| 39 |
+
"vocab_size": 50266
|
| 40 |
+
}
|
full-finetuning/german-gpt2/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.46.0"
|
| 6 |
+
}
|
full-finetuning/german-gpt2/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f8e15667a18c4b3497fb23b0fce496d6ff93c6c29494c95bc5eb396ac42a1146
|
| 3 |
+
size 497801856
|
full-finetuning/german-gpt2/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec5e28abe83c3faea24cdf4c5d768e92115d8432ee555a8db92897d89837a00c
|
| 3 |
+
size 995697594
|
full-finetuning/german-gpt2/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3bc1ac846859ad13fd356d3ed41fd20834205a7e8764e4198636863fca3f64c
|
| 3 |
+
size 14244
|
full-finetuning/german-gpt2/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a27b31e56458d82a793cfa18a3b4ed45acf96eb2c39bb55be03b452e972acf4
|
| 3 |
+
size 1064
|
full-finetuning/german-gpt2/trainer_state.json
ADDED
|
@@ -0,0 +1,1338 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.3120391368865967,
|
| 3 |
+
"best_model_checkpoint": "./models/full-finetuning/german-gpt2/checkpoint-58000",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 59835,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008356313194618534,
|
| 13 |
+
"grad_norm": 7.7337164878845215,
|
| 14 |
+
"learning_rate": 4.97e-05,
|
| 15 |
+
"loss": 5.7213,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016712626389237067,
|
| 20 |
+
"grad_norm": 6.002913951873779,
|
| 21 |
+
"learning_rate": 4.95811915395635e-05,
|
| 22 |
+
"loss": 4.8528,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.016712626389237067,
|
| 27 |
+
"eval_loss": 4.570379734039307,
|
| 28 |
+
"eval_runtime": 33.8087,
|
| 29 |
+
"eval_samples_per_second": 131.978,
|
| 30 |
+
"eval_steps_per_second": 16.505,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.025068939583855605,
|
| 35 |
+
"grad_norm": 5.462852478027344,
|
| 36 |
+
"learning_rate": 4.915985506025112e-05,
|
| 37 |
+
"loss": 4.6135,
|
| 38 |
+
"step": 1500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.033425252778474135,
|
| 42 |
+
"grad_norm": 6.940282821655273,
|
| 43 |
+
"learning_rate": 4.873851858093874e-05,
|
| 44 |
+
"loss": 4.4938,
|
| 45 |
+
"step": 2000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.033425252778474135,
|
| 49 |
+
"eval_loss": 4.2595720291137695,
|
| 50 |
+
"eval_runtime": 36.1247,
|
| 51 |
+
"eval_samples_per_second": 123.516,
|
| 52 |
+
"eval_steps_per_second": 15.446,
|
| 53 |
+
"step": 2000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.04178156597309267,
|
| 57 |
+
"grad_norm": 6.612778663635254,
|
| 58 |
+
"learning_rate": 4.8317182101626365e-05,
|
| 59 |
+
"loss": 4.3683,
|
| 60 |
+
"step": 2500
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.05013787916771121,
|
| 64 |
+
"grad_norm": 5.179454803466797,
|
| 65 |
+
"learning_rate": 4.789584562231398e-05,
|
| 66 |
+
"loss": 4.2089,
|
| 67 |
+
"step": 3000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.05013787916771121,
|
| 71 |
+
"eval_loss": 4.104772567749023,
|
| 72 |
+
"eval_runtime": 33.8319,
|
| 73 |
+
"eval_samples_per_second": 131.887,
|
| 74 |
+
"eval_steps_per_second": 16.493,
|
| 75 |
+
"step": 3000
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.05849419236232974,
|
| 79 |
+
"grad_norm": 4.007822036743164,
|
| 80 |
+
"learning_rate": 4.74745091430016e-05,
|
| 81 |
+
"loss": 4.2532,
|
| 82 |
+
"step": 3500
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.06685050555694827,
|
| 86 |
+
"grad_norm": 5.069295883178711,
|
| 87 |
+
"learning_rate": 4.7053172663689226e-05,
|
| 88 |
+
"loss": 4.2046,
|
| 89 |
+
"step": 4000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.06685050555694827,
|
| 93 |
+
"eval_loss": 4.0018157958984375,
|
| 94 |
+
"eval_runtime": 33.841,
|
| 95 |
+
"eval_samples_per_second": 131.852,
|
| 96 |
+
"eval_steps_per_second": 16.489,
|
| 97 |
+
"step": 4000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.0752068187515668,
|
| 101 |
+
"grad_norm": 8.424234390258789,
|
| 102 |
+
"learning_rate": 4.663183618437685e-05,
|
| 103 |
+
"loss": 4.1513,
|
| 104 |
+
"step": 4500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.08356313194618534,
|
| 108 |
+
"grad_norm": 5.6522016525268555,
|
| 109 |
+
"learning_rate": 4.621049970506447e-05,
|
| 110 |
+
"loss": 4.0912,
|
| 111 |
+
"step": 5000
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.08356313194618534,
|
| 115 |
+
"eval_loss": 3.927964687347412,
|
| 116 |
+
"eval_runtime": 33.8173,
|
| 117 |
+
"eval_samples_per_second": 131.944,
|
| 118 |
+
"eval_steps_per_second": 16.5,
|
| 119 |
+
"step": 5000
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.09191944514080387,
|
| 123 |
+
"grad_norm": 6.062402248382568,
|
| 124 |
+
"learning_rate": 4.5789163225752086e-05,
|
| 125 |
+
"loss": 4.0602,
|
| 126 |
+
"step": 5500
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.10027575833542242,
|
| 130 |
+
"grad_norm": 4.522783279418945,
|
| 131 |
+
"learning_rate": 4.536782674643971e-05,
|
| 132 |
+
"loss": 4.0243,
|
| 133 |
+
"step": 6000
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.10027575833542242,
|
| 137 |
+
"eval_loss": 3.8684792518615723,
|
| 138 |
+
"eval_runtime": 33.8061,
|
| 139 |
+
"eval_samples_per_second": 131.988,
|
| 140 |
+
"eval_steps_per_second": 16.506,
|
| 141 |
+
"step": 6000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.10863207153004095,
|
| 145 |
+
"grad_norm": 5.732029914855957,
|
| 146 |
+
"learning_rate": 4.494649026712733e-05,
|
| 147 |
+
"loss": 3.9523,
|
| 148 |
+
"step": 6500
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.11698838472465949,
|
| 152 |
+
"grad_norm": 5.034173011779785,
|
| 153 |
+
"learning_rate": 4.452599646077358e-05,
|
| 154 |
+
"loss": 3.9991,
|
| 155 |
+
"step": 7000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.11698838472465949,
|
| 159 |
+
"eval_loss": 3.818037748336792,
|
| 160 |
+
"eval_runtime": 33.7911,
|
| 161 |
+
"eval_samples_per_second": 132.047,
|
| 162 |
+
"eval_steps_per_second": 16.513,
|
| 163 |
+
"step": 7000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.12534469791927802,
|
| 167 |
+
"grad_norm": 6.326226711273193,
|
| 168 |
+
"learning_rate": 4.41046599814612e-05,
|
| 169 |
+
"loss": 3.9681,
|
| 170 |
+
"step": 7500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.13370101111389654,
|
| 174 |
+
"grad_norm": 5.3602495193481445,
|
| 175 |
+
"learning_rate": 4.368332350214882e-05,
|
| 176 |
+
"loss": 3.9101,
|
| 177 |
+
"step": 8000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.13370101111389654,
|
| 181 |
+
"eval_loss": 3.7742578983306885,
|
| 182 |
+
"eval_runtime": 33.7987,
|
| 183 |
+
"eval_samples_per_second": 132.017,
|
| 184 |
+
"eval_steps_per_second": 16.51,
|
| 185 |
+
"step": 8000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1420573243085151,
|
| 189 |
+
"grad_norm": 6.7508111000061035,
|
| 190 |
+
"learning_rate": 4.326198702283644e-05,
|
| 191 |
+
"loss": 3.9032,
|
| 192 |
+
"step": 8500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1504136375031336,
|
| 196 |
+
"grad_norm": 5.866630554199219,
|
| 197 |
+
"learning_rate": 4.2840650543524055e-05,
|
| 198 |
+
"loss": 3.8962,
|
| 199 |
+
"step": 9000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1504136375031336,
|
| 203 |
+
"eval_loss": 3.7523410320281982,
|
| 204 |
+
"eval_runtime": 33.8046,
|
| 205 |
+
"eval_samples_per_second": 131.994,
|
| 206 |
+
"eval_steps_per_second": 16.507,
|
| 207 |
+
"step": 9000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.15876995069775215,
|
| 211 |
+
"grad_norm": 4.693000316619873,
|
| 212 |
+
"learning_rate": 4.241931406421168e-05,
|
| 213 |
+
"loss": 3.853,
|
| 214 |
+
"step": 9500
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.16712626389237067,
|
| 218 |
+
"grad_norm": 5.653314113616943,
|
| 219 |
+
"learning_rate": 4.199882025785793e-05,
|
| 220 |
+
"loss": 3.8244,
|
| 221 |
+
"step": 10000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.16712626389237067,
|
| 225 |
+
"eval_loss": 3.711071014404297,
|
| 226 |
+
"eval_runtime": 33.8163,
|
| 227 |
+
"eval_samples_per_second": 131.948,
|
| 228 |
+
"eval_steps_per_second": 16.501,
|
| 229 |
+
"step": 10000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.17548257708698922,
|
| 233 |
+
"grad_norm": 4.951623439788818,
|
| 234 |
+
"learning_rate": 4.157748377854555e-05,
|
| 235 |
+
"loss": 3.8097,
|
| 236 |
+
"step": 10500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.18383889028160774,
|
| 240 |
+
"grad_norm": 9.94375228881836,
|
| 241 |
+
"learning_rate": 4.115614729923317e-05,
|
| 242 |
+
"loss": 3.7687,
|
| 243 |
+
"step": 11000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.18383889028160774,
|
| 247 |
+
"eval_loss": 3.6785600185394287,
|
| 248 |
+
"eval_runtime": 33.809,
|
| 249 |
+
"eval_samples_per_second": 131.977,
|
| 250 |
+
"eval_steps_per_second": 16.504,
|
| 251 |
+
"step": 11000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.1921952034762263,
|
| 255 |
+
"grad_norm": 7.155759334564209,
|
| 256 |
+
"learning_rate": 4.0734810819920794e-05,
|
| 257 |
+
"loss": 3.8008,
|
| 258 |
+
"step": 11500
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.20055151667084484,
|
| 262 |
+
"grad_norm": 3.9872701168060303,
|
| 263 |
+
"learning_rate": 4.031347434060841e-05,
|
| 264 |
+
"loss": 3.7853,
|
| 265 |
+
"step": 12000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.20055151667084484,
|
| 269 |
+
"eval_loss": 3.655622959136963,
|
| 270 |
+
"eval_runtime": 33.807,
|
| 271 |
+
"eval_samples_per_second": 131.985,
|
| 272 |
+
"eval_steps_per_second": 16.505,
|
| 273 |
+
"step": 12000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.20890782986546336,
|
| 277 |
+
"grad_norm": 7.3907647132873535,
|
| 278 |
+
"learning_rate": 3.989298053425466e-05,
|
| 279 |
+
"loss": 3.8082,
|
| 280 |
+
"step": 12500
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.2172641430600819,
|
| 284 |
+
"grad_norm": 4.089804172515869,
|
| 285 |
+
"learning_rate": 3.947164405494228e-05,
|
| 286 |
+
"loss": 3.7216,
|
| 287 |
+
"step": 13000
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.2172641430600819,
|
| 291 |
+
"eval_loss": 3.641737937927246,
|
| 292 |
+
"eval_runtime": 33.7976,
|
| 293 |
+
"eval_samples_per_second": 132.021,
|
| 294 |
+
"eval_steps_per_second": 16.51,
|
| 295 |
+
"step": 13000
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.22562045625470042,
|
| 299 |
+
"grad_norm": 3.7814671993255615,
|
| 300 |
+
"learning_rate": 3.90503075756299e-05,
|
| 301 |
+
"loss": 3.7132,
|
| 302 |
+
"step": 13500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.23397676944931897,
|
| 306 |
+
"grad_norm": 5.00490665435791,
|
| 307 |
+
"learning_rate": 3.8628971096317526e-05,
|
| 308 |
+
"loss": 3.7508,
|
| 309 |
+
"step": 14000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.23397676944931897,
|
| 313 |
+
"eval_loss": 3.6097123622894287,
|
| 314 |
+
"eval_runtime": 33.8097,
|
| 315 |
+
"eval_samples_per_second": 131.974,
|
| 316 |
+
"eval_steps_per_second": 16.504,
|
| 317 |
+
"step": 14000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.2423330826439375,
|
| 321 |
+
"grad_norm": 4.1177659034729,
|
| 322 |
+
"learning_rate": 3.820763461700515e-05,
|
| 323 |
+
"loss": 3.7317,
|
| 324 |
+
"step": 14500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.25068939583855604,
|
| 328 |
+
"grad_norm": 5.807891845703125,
|
| 329 |
+
"learning_rate": 3.7786298137692763e-05,
|
| 330 |
+
"loss": 3.7186,
|
| 331 |
+
"step": 15000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.25068939583855604,
|
| 335 |
+
"eval_loss": 3.5940768718719482,
|
| 336 |
+
"eval_runtime": 33.8398,
|
| 337 |
+
"eval_samples_per_second": 131.856,
|
| 338 |
+
"eval_steps_per_second": 16.489,
|
| 339 |
+
"step": 15000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2590457090331746,
|
| 343 |
+
"grad_norm": 5.580984592437744,
|
| 344 |
+
"learning_rate": 3.736496165838038e-05,
|
| 345 |
+
"loss": 3.7059,
|
| 346 |
+
"step": 15500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2674020222277931,
|
| 350 |
+
"grad_norm": 5.553014278411865,
|
| 351 |
+
"learning_rate": 3.6943625179068e-05,
|
| 352 |
+
"loss": 3.6317,
|
| 353 |
+
"step": 16000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2674020222277931,
|
| 357 |
+
"eval_loss": 3.5815696716308594,
|
| 358 |
+
"eval_runtime": 33.7945,
|
| 359 |
+
"eval_samples_per_second": 132.033,
|
| 360 |
+
"eval_steps_per_second": 16.512,
|
| 361 |
+
"step": 16000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.2757583354224116,
|
| 365 |
+
"grad_norm": 7.818964958190918,
|
| 366 |
+
"learning_rate": 3.652313137271425e-05,
|
| 367 |
+
"loss": 3.6837,
|
| 368 |
+
"step": 16500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2841146486170302,
|
| 372 |
+
"grad_norm": 7.232082843780518,
|
| 373 |
+
"learning_rate": 3.610179489340187e-05,
|
| 374 |
+
"loss": 3.6965,
|
| 375 |
+
"step": 17000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.2841146486170302,
|
| 379 |
+
"eval_loss": 3.559220552444458,
|
| 380 |
+
"eval_runtime": 33.8032,
|
| 381 |
+
"eval_samples_per_second": 131.999,
|
| 382 |
+
"eval_steps_per_second": 16.507,
|
| 383 |
+
"step": 17000
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.2924709618116487,
|
| 387 |
+
"grad_norm": 5.058260917663574,
|
| 388 |
+
"learning_rate": 3.5680458414089495e-05,
|
| 389 |
+
"loss": 3.6218,
|
| 390 |
+
"step": 17500
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.3008272750062672,
|
| 394 |
+
"grad_norm": 6.645322322845459,
|
| 395 |
+
"learning_rate": 3.525912193477712e-05,
|
| 396 |
+
"loss": 3.6203,
|
| 397 |
+
"step": 18000
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.3008272750062672,
|
| 401 |
+
"eval_loss": 3.5445735454559326,
|
| 402 |
+
"eval_runtime": 33.826,
|
| 403 |
+
"eval_samples_per_second": 131.911,
|
| 404 |
+
"eval_steps_per_second": 16.496,
|
| 405 |
+
"step": 18000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.30918358820088576,
|
| 409 |
+
"grad_norm": 6.7204084396362305,
|
| 410 |
+
"learning_rate": 3.483862812842336e-05,
|
| 411 |
+
"loss": 3.6135,
|
| 412 |
+
"step": 18500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.3175399013955043,
|
| 416 |
+
"grad_norm": 5.846601486206055,
|
| 417 |
+
"learning_rate": 3.441729164911098e-05,
|
| 418 |
+
"loss": 3.6139,
|
| 419 |
+
"step": 19000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.3175399013955043,
|
| 423 |
+
"eval_loss": 3.529482364654541,
|
| 424 |
+
"eval_runtime": 33.8084,
|
| 425 |
+
"eval_samples_per_second": 131.979,
|
| 426 |
+
"eval_steps_per_second": 16.505,
|
| 427 |
+
"step": 19000
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.32589621459012286,
|
| 431 |
+
"grad_norm": 4.412099838256836,
|
| 432 |
+
"learning_rate": 3.3995955169798604e-05,
|
| 433 |
+
"loss": 3.6202,
|
| 434 |
+
"step": 19500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.33425252778474135,
|
| 438 |
+
"grad_norm": 6.884310722351074,
|
| 439 |
+
"learning_rate": 3.357461869048623e-05,
|
| 440 |
+
"loss": 3.5862,
|
| 441 |
+
"step": 20000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.33425252778474135,
|
| 445 |
+
"eval_loss": 3.518963098526001,
|
| 446 |
+
"eval_runtime": 33.8217,
|
| 447 |
+
"eval_samples_per_second": 131.927,
|
| 448 |
+
"eval_steps_per_second": 16.498,
|
| 449 |
+
"step": 20000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.3426088409793599,
|
| 453 |
+
"grad_norm": 4.762236595153809,
|
| 454 |
+
"learning_rate": 3.315328221117385e-05,
|
| 455 |
+
"loss": 3.613,
|
| 456 |
+
"step": 20500
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.35096515417397844,
|
| 460 |
+
"grad_norm": 3.5604190826416016,
|
| 461 |
+
"learning_rate": 3.273278840482009e-05,
|
| 462 |
+
"loss": 3.5767,
|
| 463 |
+
"step": 21000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.35096515417397844,
|
| 467 |
+
"eval_loss": 3.5067789554595947,
|
| 468 |
+
"eval_runtime": 33.8036,
|
| 469 |
+
"eval_samples_per_second": 131.998,
|
| 470 |
+
"eval_steps_per_second": 16.507,
|
| 471 |
+
"step": 21000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.359321467368597,
|
| 475 |
+
"grad_norm": 3.5096914768218994,
|
| 476 |
+
"learning_rate": 3.2311451925507714e-05,
|
| 477 |
+
"loss": 3.5692,
|
| 478 |
+
"step": 21500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.3676777805632155,
|
| 482 |
+
"grad_norm": 4.54230260848999,
|
| 483 |
+
"learning_rate": 3.1890115446195336e-05,
|
| 484 |
+
"loss": 3.5914,
|
| 485 |
+
"step": 22000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.3676777805632155,
|
| 489 |
+
"eval_loss": 3.491032361984253,
|
| 490 |
+
"eval_runtime": 33.8108,
|
| 491 |
+
"eval_samples_per_second": 131.97,
|
| 492 |
+
"eval_steps_per_second": 16.504,
|
| 493 |
+
"step": 22000
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.37603409375783403,
|
| 497 |
+
"grad_norm": 5.874061584472656,
|
| 498 |
+
"learning_rate": 3.146877896688296e-05,
|
| 499 |
+
"loss": 3.6494,
|
| 500 |
+
"step": 22500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3843904069524526,
|
| 504 |
+
"grad_norm": 5.265365123748779,
|
| 505 |
+
"learning_rate": 3.10482851605292e-05,
|
| 506 |
+
"loss": 3.5796,
|
| 507 |
+
"step": 23000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.3843904069524526,
|
| 511 |
+
"eval_loss": 3.4834258556365967,
|
| 512 |
+
"eval_runtime": 33.7961,
|
| 513 |
+
"eval_samples_per_second": 132.027,
|
| 514 |
+
"eval_steps_per_second": 16.511,
|
| 515 |
+
"step": 23000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.3927467201470711,
|
| 519 |
+
"grad_norm": 4.7764129638671875,
|
| 520 |
+
"learning_rate": 3.062694868121682e-05,
|
| 521 |
+
"loss": 3.5603,
|
| 522 |
+
"step": 23500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.4011030333416897,
|
| 526 |
+
"grad_norm": 4.89961576461792,
|
| 527 |
+
"learning_rate": 3.0205612201904442e-05,
|
| 528 |
+
"loss": 3.5427,
|
| 529 |
+
"step": 24000
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.4011030333416897,
|
| 533 |
+
"eval_loss": 3.476203203201294,
|
| 534 |
+
"eval_runtime": 33.8616,
|
| 535 |
+
"eval_samples_per_second": 131.772,
|
| 536 |
+
"eval_steps_per_second": 16.479,
|
| 537 |
+
"step": 24000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.40945934653630817,
|
| 541 |
+
"grad_norm": 6.659942150115967,
|
| 542 |
+
"learning_rate": 2.9784275722592064e-05,
|
| 543 |
+
"loss": 3.5559,
|
| 544 |
+
"step": 24500
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.4178156597309267,
|
| 548 |
+
"grad_norm": 4.591799259185791,
|
| 549 |
+
"learning_rate": 2.936378191623831e-05,
|
| 550 |
+
"loss": 3.5087,
|
| 551 |
+
"step": 25000
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.4178156597309267,
|
| 555 |
+
"eval_loss": 3.465503454208374,
|
| 556 |
+
"eval_runtime": 33.8191,
|
| 557 |
+
"eval_samples_per_second": 131.937,
|
| 558 |
+
"eval_steps_per_second": 16.5,
|
| 559 |
+
"step": 25000
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.42617197292554526,
|
| 563 |
+
"grad_norm": 4.84548282623291,
|
| 564 |
+
"learning_rate": 2.894244543692593e-05,
|
| 565 |
+
"loss": 3.5474,
|
| 566 |
+
"step": 25500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.4345282861201638,
|
| 570 |
+
"grad_norm": 5.839838027954102,
|
| 571 |
+
"learning_rate": 2.852110895761355e-05,
|
| 572 |
+
"loss": 3.5723,
|
| 573 |
+
"step": 26000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.4345282861201638,
|
| 577 |
+
"eval_loss": 3.458843946456909,
|
| 578 |
+
"eval_runtime": 33.8846,
|
| 579 |
+
"eval_samples_per_second": 131.682,
|
| 580 |
+
"eval_steps_per_second": 16.468,
|
| 581 |
+
"step": 26000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.4428845993147823,
|
| 585 |
+
"grad_norm": 3.7624571323394775,
|
| 586 |
+
"learning_rate": 2.8099772478301174e-05,
|
| 587 |
+
"loss": 3.4898,
|
| 588 |
+
"step": 26500
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.45124091250940085,
|
| 592 |
+
"grad_norm": 4.763157844543457,
|
| 593 |
+
"learning_rate": 2.7678435998988793e-05,
|
| 594 |
+
"loss": 3.5493,
|
| 595 |
+
"step": 27000
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.45124091250940085,
|
| 599 |
+
"eval_loss": 3.4483096599578857,
|
| 600 |
+
"eval_runtime": 33.7859,
|
| 601 |
+
"eval_samples_per_second": 132.067,
|
| 602 |
+
"eval_steps_per_second": 16.516,
|
| 603 |
+
"step": 27000
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.4595972257040194,
|
| 607 |
+
"grad_norm": 4.820297718048096,
|
| 608 |
+
"learning_rate": 2.725794219263504e-05,
|
| 609 |
+
"loss": 3.5162,
|
| 610 |
+
"step": 27500
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.46795353889863794,
|
| 614 |
+
"grad_norm": 8.342415809631348,
|
| 615 |
+
"learning_rate": 2.683660571332266e-05,
|
| 616 |
+
"loss": 3.4815,
|
| 617 |
+
"step": 28000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.46795353889863794,
|
| 621 |
+
"eval_loss": 3.4425435066223145,
|
| 622 |
+
"eval_runtime": 33.8093,
|
| 623 |
+
"eval_samples_per_second": 131.976,
|
| 624 |
+
"eval_steps_per_second": 16.504,
|
| 625 |
+
"step": 28000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.47630985209325644,
|
| 629 |
+
"grad_norm": 3.9816768169403076,
|
| 630 |
+
"learning_rate": 2.641526923401028e-05,
|
| 631 |
+
"loss": 3.5168,
|
| 632 |
+
"step": 28500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.484666165287875,
|
| 636 |
+
"grad_norm": 4.662781715393066,
|
| 637 |
+
"learning_rate": 2.5993932754697902e-05,
|
| 638 |
+
"loss": 3.5442,
|
| 639 |
+
"step": 29000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.484666165287875,
|
| 643 |
+
"eval_loss": 3.4301340579986572,
|
| 644 |
+
"eval_runtime": 33.7983,
|
| 645 |
+
"eval_samples_per_second": 132.018,
|
| 646 |
+
"eval_steps_per_second": 16.51,
|
| 647 |
+
"step": 29000
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.49302247848249353,
|
| 651 |
+
"grad_norm": 5.855069160461426,
|
| 652 |
+
"learning_rate": 2.5573438948344148e-05,
|
| 653 |
+
"loss": 3.5048,
|
| 654 |
+
"step": 29500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5013787916771121,
|
| 658 |
+
"grad_norm": 3.8957674503326416,
|
| 659 |
+
"learning_rate": 2.515210246903177e-05,
|
| 660 |
+
"loss": 3.5452,
|
| 661 |
+
"step": 30000
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5013787916771121,
|
| 665 |
+
"eval_loss": 3.4219019412994385,
|
| 666 |
+
"eval_runtime": 33.8062,
|
| 667 |
+
"eval_samples_per_second": 131.988,
|
| 668 |
+
"eval_steps_per_second": 16.506,
|
| 669 |
+
"step": 30000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.5097351048717306,
|
| 673 |
+
"grad_norm": 5.273784160614014,
|
| 674 |
+
"learning_rate": 2.473076598971939e-05,
|
| 675 |
+
"loss": 3.4951,
|
| 676 |
+
"step": 30500
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.5180914180663492,
|
| 680 |
+
"grad_norm": 6.109909534454346,
|
| 681 |
+
"learning_rate": 2.430942951040701e-05,
|
| 682 |
+
"loss": 3.4879,
|
| 683 |
+
"step": 31000
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.5180914180663492,
|
| 687 |
+
"eval_loss": 3.419405698776245,
|
| 688 |
+
"eval_runtime": 33.7875,
|
| 689 |
+
"eval_samples_per_second": 132.061,
|
| 690 |
+
"eval_steps_per_second": 16.515,
|
| 691 |
+
"step": 31000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.5264477312609677,
|
| 695 |
+
"grad_norm": 7.194368839263916,
|
| 696 |
+
"learning_rate": 2.3888935704053257e-05,
|
| 697 |
+
"loss": 3.5259,
|
| 698 |
+
"step": 31500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.5348040444555862,
|
| 702 |
+
"grad_norm": 4.303890228271484,
|
| 703 |
+
"learning_rate": 2.3468441897699503e-05,
|
| 704 |
+
"loss": 3.4956,
|
| 705 |
+
"step": 32000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.5348040444555862,
|
| 709 |
+
"eval_loss": 3.4062435626983643,
|
| 710 |
+
"eval_runtime": 33.8183,
|
| 711 |
+
"eval_samples_per_second": 131.94,
|
| 712 |
+
"eval_steps_per_second": 16.5,
|
| 713 |
+
"step": 32000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.5431603576502048,
|
| 717 |
+
"grad_norm": 4.180033206939697,
|
| 718 |
+
"learning_rate": 2.3047105418387125e-05,
|
| 719 |
+
"loss": 3.5137,
|
| 720 |
+
"step": 32500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.5515166708448233,
|
| 724 |
+
"grad_norm": 4.65141487121582,
|
| 725 |
+
"learning_rate": 2.2625768939074744e-05,
|
| 726 |
+
"loss": 3.4635,
|
| 727 |
+
"step": 33000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.5515166708448233,
|
| 731 |
+
"eval_loss": 3.3979876041412354,
|
| 732 |
+
"eval_runtime": 33.8238,
|
| 733 |
+
"eval_samples_per_second": 131.919,
|
| 734 |
+
"eval_steps_per_second": 16.497,
|
| 735 |
+
"step": 33000
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.5598729840394417,
|
| 739 |
+
"grad_norm": 7.3677215576171875,
|
| 740 |
+
"learning_rate": 2.2204432459762367e-05,
|
| 741 |
+
"loss": 3.5071,
|
| 742 |
+
"step": 33500
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.5682292972340603,
|
| 746 |
+
"grad_norm": 4.072124481201172,
|
| 747 |
+
"learning_rate": 2.178309598044999e-05,
|
| 748 |
+
"loss": 3.4836,
|
| 749 |
+
"step": 34000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.5682292972340603,
|
| 753 |
+
"eval_loss": 3.3961925506591797,
|
| 754 |
+
"eval_runtime": 33.8074,
|
| 755 |
+
"eval_samples_per_second": 131.983,
|
| 756 |
+
"eval_steps_per_second": 16.505,
|
| 757 |
+
"step": 34000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.5765856104286788,
|
| 761 |
+
"grad_norm": 4.479409217834473,
|
| 762 |
+
"learning_rate": 2.136175950113761e-05,
|
| 763 |
+
"loss": 3.4763,
|
| 764 |
+
"step": 34500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.5849419236232974,
|
| 768 |
+
"grad_norm": 3.3214428424835205,
|
| 769 |
+
"learning_rate": 2.094042302182523e-05,
|
| 770 |
+
"loss": 3.4378,
|
| 771 |
+
"step": 35000
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.5849419236232974,
|
| 775 |
+
"eval_loss": 3.389758825302124,
|
| 776 |
+
"eval_runtime": 33.7986,
|
| 777 |
+
"eval_samples_per_second": 132.017,
|
| 778 |
+
"eval_steps_per_second": 16.51,
|
| 779 |
+
"step": 35000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5932982368179159,
|
| 783 |
+
"grad_norm": 6.104400157928467,
|
| 784 |
+
"learning_rate": 2.0519086542512852e-05,
|
| 785 |
+
"loss": 3.4959,
|
| 786 |
+
"step": 35500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.6016545500125344,
|
| 790 |
+
"grad_norm": 5.734145641326904,
|
| 791 |
+
"learning_rate": 2.0097750063200475e-05,
|
| 792 |
+
"loss": 3.4673,
|
| 793 |
+
"step": 36000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6016545500125344,
|
| 797 |
+
"eval_loss": 3.3848955631256104,
|
| 798 |
+
"eval_runtime": 33.8025,
|
| 799 |
+
"eval_samples_per_second": 132.002,
|
| 800 |
+
"eval_steps_per_second": 16.508,
|
| 801 |
+
"step": 36000
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.610010863207153,
|
| 805 |
+
"grad_norm": 4.3256683349609375,
|
| 806 |
+
"learning_rate": 1.9677256256846717e-05,
|
| 807 |
+
"loss": 3.4903,
|
| 808 |
+
"step": 36500
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6183671764017715,
|
| 812 |
+
"grad_norm": 5.769290447235107,
|
| 813 |
+
"learning_rate": 1.925591977753434e-05,
|
| 814 |
+
"loss": 3.4722,
|
| 815 |
+
"step": 37000
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.6183671764017715,
|
| 819 |
+
"eval_loss": 3.3791496753692627,
|
| 820 |
+
"eval_runtime": 33.8066,
|
| 821 |
+
"eval_samples_per_second": 131.986,
|
| 822 |
+
"eval_steps_per_second": 16.506,
|
| 823 |
+
"step": 37000
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.6267234895963901,
|
| 827 |
+
"grad_norm": 4.285965442657471,
|
| 828 |
+
"learning_rate": 1.8834583298221962e-05,
|
| 829 |
+
"loss": 3.4922,
|
| 830 |
+
"step": 37500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.6350798027910086,
|
| 834 |
+
"grad_norm": 4.97435998916626,
|
| 835 |
+
"learning_rate": 1.8413246818909584e-05,
|
| 836 |
+
"loss": 3.4814,
|
| 837 |
+
"step": 38000
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.6350798027910086,
|
| 841 |
+
"eval_loss": 3.3762271404266357,
|
| 842 |
+
"eval_runtime": 33.8323,
|
| 843 |
+
"eval_samples_per_second": 131.886,
|
| 844 |
+
"eval_steps_per_second": 16.493,
|
| 845 |
+
"step": 38000
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.6434361159856271,
|
| 849 |
+
"grad_norm": 4.531788349151611,
|
| 850 |
+
"learning_rate": 1.7992753012555827e-05,
|
| 851 |
+
"loss": 3.4144,
|
| 852 |
+
"step": 38500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.6517924291802457,
|
| 856 |
+
"grad_norm": 5.197137355804443,
|
| 857 |
+
"learning_rate": 1.757141653324345e-05,
|
| 858 |
+
"loss": 3.4358,
|
| 859 |
+
"step": 39000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.6517924291802457,
|
| 863 |
+
"eval_loss": 3.3685717582702637,
|
| 864 |
+
"eval_runtime": 33.7899,
|
| 865 |
+
"eval_samples_per_second": 132.051,
|
| 866 |
+
"eval_steps_per_second": 16.514,
|
| 867 |
+
"step": 39000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.6601487423748642,
|
| 871 |
+
"grad_norm": 5.673033714294434,
|
| 872 |
+
"learning_rate": 1.715008005393107e-05,
|
| 873 |
+
"loss": 3.3936,
|
| 874 |
+
"step": 39500
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.6685050555694827,
|
| 878 |
+
"grad_norm": 4.699774742126465,
|
| 879 |
+
"learning_rate": 1.672874357461869e-05,
|
| 880 |
+
"loss": 3.4711,
|
| 881 |
+
"step": 40000
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.6685050555694827,
|
| 885 |
+
"eval_loss": 3.3646111488342285,
|
| 886 |
+
"eval_runtime": 33.8073,
|
| 887 |
+
"eval_samples_per_second": 131.983,
|
| 888 |
+
"eval_steps_per_second": 16.505,
|
| 889 |
+
"step": 40000
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.6768613687641013,
|
| 893 |
+
"grad_norm": 4.5016889572143555,
|
| 894 |
+
"learning_rate": 1.630824976826494e-05,
|
| 895 |
+
"loss": 3.4073,
|
| 896 |
+
"step": 40500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.6852176819587198,
|
| 900 |
+
"grad_norm": 9.633309364318848,
|
| 901 |
+
"learning_rate": 1.5886913288952558e-05,
|
| 902 |
+
"loss": 3.4437,
|
| 903 |
+
"step": 41000
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.6852176819587198,
|
| 907 |
+
"eval_loss": 3.3601598739624023,
|
| 908 |
+
"eval_runtime": 33.7786,
|
| 909 |
+
"eval_samples_per_second": 132.096,
|
| 910 |
+
"eval_steps_per_second": 16.519,
|
| 911 |
+
"step": 41000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.6935739951533384,
|
| 915 |
+
"grad_norm": 3.356966733932495,
|
| 916 |
+
"learning_rate": 1.5465576809640177e-05,
|
| 917 |
+
"loss": 3.3706,
|
| 918 |
+
"step": 41500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.7019303083479569,
|
| 922 |
+
"grad_norm": 4.3639678955078125,
|
| 923 |
+
"learning_rate": 1.50442403303278e-05,
|
| 924 |
+
"loss": 3.4171,
|
| 925 |
+
"step": 42000
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.7019303083479569,
|
| 929 |
+
"eval_loss": 3.358569622039795,
|
| 930 |
+
"eval_runtime": 33.8063,
|
| 931 |
+
"eval_samples_per_second": 131.987,
|
| 932 |
+
"eval_steps_per_second": 16.506,
|
| 933 |
+
"step": 42000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.7102866215425754,
|
| 937 |
+
"grad_norm": 6.138998985290527,
|
| 938 |
+
"learning_rate": 1.4623746523974047e-05,
|
| 939 |
+
"loss": 3.3822,
|
| 940 |
+
"step": 42500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.718642934737194,
|
| 944 |
+
"grad_norm": 5.975950717926025,
|
| 945 |
+
"learning_rate": 1.4202410044661668e-05,
|
| 946 |
+
"loss": 3.442,
|
| 947 |
+
"step": 43000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.718642934737194,
|
| 951 |
+
"eval_loss": 3.3521432876586914,
|
| 952 |
+
"eval_runtime": 33.7743,
|
| 953 |
+
"eval_samples_per_second": 132.112,
|
| 954 |
+
"eval_steps_per_second": 16.521,
|
| 955 |
+
"step": 43000
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.7269992479318125,
|
| 959 |
+
"grad_norm": 5.316736698150635,
|
| 960 |
+
"learning_rate": 1.378107356534929e-05,
|
| 961 |
+
"loss": 3.3837,
|
| 962 |
+
"step": 43500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.735355561126431,
|
| 966 |
+
"grad_norm": 4.64693021774292,
|
| 967 |
+
"learning_rate": 1.335973708603691e-05,
|
| 968 |
+
"loss": 3.3846,
|
| 969 |
+
"step": 44000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.735355561126431,
|
| 973 |
+
"eval_loss": 3.3473236560821533,
|
| 974 |
+
"eval_runtime": 33.8058,
|
| 975 |
+
"eval_samples_per_second": 131.989,
|
| 976 |
+
"eval_steps_per_second": 16.506,
|
| 977 |
+
"step": 44000
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.7437118743210496,
|
| 981 |
+
"grad_norm": 7.507499694824219,
|
| 982 |
+
"learning_rate": 1.293840060672453e-05,
|
| 983 |
+
"loss": 3.4229,
|
| 984 |
+
"step": 44500
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.7520681875156681,
|
| 988 |
+
"grad_norm": 4.695137023925781,
|
| 989 |
+
"learning_rate": 1.2517906800370777e-05,
|
| 990 |
+
"loss": 3.4064,
|
| 991 |
+
"step": 45000
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.7520681875156681,
|
| 995 |
+
"eval_loss": 3.3478496074676514,
|
| 996 |
+
"eval_runtime": 33.7907,
|
| 997 |
+
"eval_samples_per_second": 132.048,
|
| 998 |
+
"eval_steps_per_second": 16.513,
|
| 999 |
+
"step": 45000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.7604245007102867,
|
| 1003 |
+
"grad_norm": 6.782580375671387,
|
| 1004 |
+
"learning_rate": 1.2096570321058398e-05,
|
| 1005 |
+
"loss": 3.4665,
|
| 1006 |
+
"step": 45500
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.7687808139049052,
|
| 1010 |
+
"grad_norm": 7.160044193267822,
|
| 1011 |
+
"learning_rate": 1.167523384174602e-05,
|
| 1012 |
+
"loss": 3.4181,
|
| 1013 |
+
"step": 46000
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.7687808139049052,
|
| 1017 |
+
"eval_loss": 3.339061737060547,
|
| 1018 |
+
"eval_runtime": 33.8357,
|
| 1019 |
+
"eval_samples_per_second": 131.872,
|
| 1020 |
+
"eval_steps_per_second": 16.491,
|
| 1021 |
+
"step": 46000
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 0.7771371270995237,
|
| 1025 |
+
"grad_norm": 6.559309482574463,
|
| 1026 |
+
"learning_rate": 1.1253897362433639e-05,
|
| 1027 |
+
"loss": 3.3691,
|
| 1028 |
+
"step": 46500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.7854934402941423,
|
| 1032 |
+
"grad_norm": 4.2476043701171875,
|
| 1033 |
+
"learning_rate": 1.0832560883121261e-05,
|
| 1034 |
+
"loss": 3.3825,
|
| 1035 |
+
"step": 47000
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.7854934402941423,
|
| 1039 |
+
"eval_loss": 3.3343887329101562,
|
| 1040 |
+
"eval_runtime": 33.7813,
|
| 1041 |
+
"eval_samples_per_second": 132.085,
|
| 1042 |
+
"eval_steps_per_second": 16.518,
|
| 1043 |
+
"step": 47000
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.7938497534887607,
|
| 1047 |
+
"grad_norm": 8.163984298706055,
|
| 1048 |
+
"learning_rate": 1.0411224403808882e-05,
|
| 1049 |
+
"loss": 3.4073,
|
| 1050 |
+
"step": 47500
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.8022060666833793,
|
| 1054 |
+
"grad_norm": 4.705440998077393,
|
| 1055 |
+
"learning_rate": 9.989887924496504e-06,
|
| 1056 |
+
"loss": 3.4277,
|
| 1057 |
+
"step": 48000
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.8022060666833793,
|
| 1061 |
+
"eval_loss": 3.334386110305786,
|
| 1062 |
+
"eval_runtime": 33.8549,
|
| 1063 |
+
"eval_samples_per_second": 131.798,
|
| 1064 |
+
"eval_steps_per_second": 16.482,
|
| 1065 |
+
"step": 48000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.8105623798779978,
|
| 1069 |
+
"grad_norm": 5.080317497253418,
|
| 1070 |
+
"learning_rate": 9.56939411814275e-06,
|
| 1071 |
+
"loss": 3.4014,
|
| 1072 |
+
"step": 48500
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.8189186930726163,
|
| 1076 |
+
"grad_norm": 4.523305892944336,
|
| 1077 |
+
"learning_rate": 9.14805763883037e-06,
|
| 1078 |
+
"loss": 3.331,
|
| 1079 |
+
"step": 49000
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.8189186930726163,
|
| 1083 |
+
"eval_loss": 3.33245587348938,
|
| 1084 |
+
"eval_runtime": 33.9406,
|
| 1085 |
+
"eval_samples_per_second": 131.465,
|
| 1086 |
+
"eval_steps_per_second": 16.44,
|
| 1087 |
+
"step": 49000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.8272750062672349,
|
| 1091 |
+
"grad_norm": 5.878600597381592,
|
| 1092 |
+
"learning_rate": 8.726721159517993e-06,
|
| 1093 |
+
"loss": 3.4097,
|
| 1094 |
+
"step": 49500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.8356313194618534,
|
| 1098 |
+
"grad_norm": 4.23352575302124,
|
| 1099 |
+
"learning_rate": 8.305384680205612e-06,
|
| 1100 |
+
"loss": 3.4073,
|
| 1101 |
+
"step": 50000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.8356313194618534,
|
| 1105 |
+
"eval_loss": 3.328063488006592,
|
| 1106 |
+
"eval_runtime": 33.9523,
|
| 1107 |
+
"eval_samples_per_second": 131.42,
|
| 1108 |
+
"eval_steps_per_second": 16.435,
|
| 1109 |
+
"step": 50000
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.8439876326564719,
|
| 1113 |
+
"grad_norm": 6.142092227935791,
|
| 1114 |
+
"learning_rate": 7.884890873851858e-06,
|
| 1115 |
+
"loss": 3.3904,
|
| 1116 |
+
"step": 50500
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.8523439458510905,
|
| 1120 |
+
"grad_norm": 7.910597324371338,
|
| 1121 |
+
"learning_rate": 7.463554394539479e-06,
|
| 1122 |
+
"loss": 3.3741,
|
| 1123 |
+
"step": 51000
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.8523439458510905,
|
| 1127 |
+
"eval_loss": 3.327683448791504,
|
| 1128 |
+
"eval_runtime": 33.9417,
|
| 1129 |
+
"eval_samples_per_second": 131.461,
|
| 1130 |
+
"eval_steps_per_second": 16.44,
|
| 1131 |
+
"step": 51000
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.860700259045709,
|
| 1135 |
+
"grad_norm": 4.868673324584961,
|
| 1136 |
+
"learning_rate": 7.0422179152271005e-06,
|
| 1137 |
+
"loss": 3.3615,
|
| 1138 |
+
"step": 51500
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.8690565722403276,
|
| 1142 |
+
"grad_norm": 4.47282075881958,
|
| 1143 |
+
"learning_rate": 6.620881435914722e-06,
|
| 1144 |
+
"loss": 3.3467,
|
| 1145 |
+
"step": 52000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.8690565722403276,
|
| 1149 |
+
"eval_loss": 3.3206417560577393,
|
| 1150 |
+
"eval_runtime": 33.9544,
|
| 1151 |
+
"eval_samples_per_second": 131.412,
|
| 1152 |
+
"eval_steps_per_second": 16.434,
|
| 1153 |
+
"step": 52000
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.8774128854349461,
|
| 1157 |
+
"grad_norm": 5.413076877593994,
|
| 1158 |
+
"learning_rate": 6.200387629560968e-06,
|
| 1159 |
+
"loss": 3.3221,
|
| 1160 |
+
"step": 52500
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.8857691986295646,
|
| 1164 |
+
"grad_norm": 5.672983169555664,
|
| 1165 |
+
"learning_rate": 5.779051150248588e-06,
|
| 1166 |
+
"loss": 3.418,
|
| 1167 |
+
"step": 53000
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.8857691986295646,
|
| 1171 |
+
"eval_loss": 3.320605516433716,
|
| 1172 |
+
"eval_runtime": 33.884,
|
| 1173 |
+
"eval_samples_per_second": 131.684,
|
| 1174 |
+
"eval_steps_per_second": 16.468,
|
| 1175 |
+
"step": 53000
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.8941255118241832,
|
| 1179 |
+
"grad_norm": 8.368507385253906,
|
| 1180 |
+
"learning_rate": 5.35771467093621e-06,
|
| 1181 |
+
"loss": 3.3747,
|
| 1182 |
+
"step": 53500
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.9024818250188017,
|
| 1186 |
+
"grad_norm": 6.918625831604004,
|
| 1187 |
+
"learning_rate": 4.936378191623831e-06,
|
| 1188 |
+
"loss": 3.416,
|
| 1189 |
+
"step": 54000
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.9024818250188017,
|
| 1193 |
+
"eval_loss": 3.317139148712158,
|
| 1194 |
+
"eval_runtime": 33.9331,
|
| 1195 |
+
"eval_samples_per_second": 131.494,
|
| 1196 |
+
"eval_steps_per_second": 16.444,
|
| 1197 |
+
"step": 54000
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.9108381382134202,
|
| 1201 |
+
"grad_norm": 6.566315174102783,
|
| 1202 |
+
"learning_rate": 4.515884385270077e-06,
|
| 1203 |
+
"loss": 3.3985,
|
| 1204 |
+
"step": 54500
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 0.9191944514080388,
|
| 1208 |
+
"grad_norm": 4.790822982788086,
|
| 1209 |
+
"learning_rate": 4.094547905957698e-06,
|
| 1210 |
+
"loss": 3.3607,
|
| 1211 |
+
"step": 55000
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.9191944514080388,
|
| 1215 |
+
"eval_loss": 3.3170626163482666,
|
| 1216 |
+
"eval_runtime": 33.9426,
|
| 1217 |
+
"eval_samples_per_second": 131.457,
|
| 1218 |
+
"eval_steps_per_second": 16.439,
|
| 1219 |
+
"step": 55000
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 0.9275507646026573,
|
| 1223 |
+
"grad_norm": 5.501828193664551,
|
| 1224 |
+
"learning_rate": 3.6732114266453192e-06,
|
| 1225 |
+
"loss": 3.3586,
|
| 1226 |
+
"step": 55500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.9359070777972759,
|
| 1230 |
+
"grad_norm": 8.011107444763184,
|
| 1231 |
+
"learning_rate": 3.2518749473329403e-06,
|
| 1232 |
+
"loss": 3.4076,
|
| 1233 |
+
"step": 56000
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.9359070777972759,
|
| 1237 |
+
"eval_loss": 3.3138890266418457,
|
| 1238 |
+
"eval_runtime": 33.9309,
|
| 1239 |
+
"eval_samples_per_second": 131.503,
|
| 1240 |
+
"eval_steps_per_second": 16.445,
|
| 1241 |
+
"step": 56000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.9442633909918944,
|
| 1245 |
+
"grad_norm": 5.630584239959717,
|
| 1246 |
+
"learning_rate": 2.8305384680205613e-06,
|
| 1247 |
+
"loss": 3.4309,
|
| 1248 |
+
"step": 56500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.9526197041865129,
|
| 1252 |
+
"grad_norm": 3.990445137023926,
|
| 1253 |
+
"learning_rate": 2.4092019887081824e-06,
|
| 1254 |
+
"loss": 3.3257,
|
| 1255 |
+
"step": 57000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.9526197041865129,
|
| 1259 |
+
"eval_loss": 3.3135273456573486,
|
| 1260 |
+
"eval_runtime": 33.8948,
|
| 1261 |
+
"eval_samples_per_second": 131.643,
|
| 1262 |
+
"eval_steps_per_second": 16.463,
|
| 1263 |
+
"step": 57000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.9609760173811315,
|
| 1267 |
+
"grad_norm": 4.2760396003723145,
|
| 1268 |
+
"learning_rate": 1.9878655093958034e-06,
|
| 1269 |
+
"loss": 3.3362,
|
| 1270 |
+
"step": 57500
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.96933233057575,
|
| 1274 |
+
"grad_norm": 4.5603532791137695,
|
| 1275 |
+
"learning_rate": 1.5665290300834245e-06,
|
| 1276 |
+
"loss": 3.3657,
|
| 1277 |
+
"step": 58000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.96933233057575,
|
| 1281 |
+
"eval_loss": 3.3120391368865967,
|
| 1282 |
+
"eval_runtime": 33.957,
|
| 1283 |
+
"eval_samples_per_second": 131.401,
|
| 1284 |
+
"eval_steps_per_second": 16.433,
|
| 1285 |
+
"step": 58000
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.9776886437703685,
|
| 1289 |
+
"grad_norm": 3.6152896881103516,
|
| 1290 |
+
"learning_rate": 1.1460352237296705e-06,
|
| 1291 |
+
"loss": 3.3847,
|
| 1292 |
+
"step": 58500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.9860449569649871,
|
| 1296 |
+
"grad_norm": 5.7242751121521,
|
| 1297 |
+
"learning_rate": 7.246987444172916e-07,
|
| 1298 |
+
"loss": 3.3509,
|
| 1299 |
+
"step": 59000
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.9860449569649871,
|
| 1303 |
+
"eval_loss": 3.3113863468170166,
|
| 1304 |
+
"eval_runtime": 33.9749,
|
| 1305 |
+
"eval_samples_per_second": 131.332,
|
| 1306 |
+
"eval_steps_per_second": 16.424,
|
| 1307 |
+
"step": 59000
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.9944012701596056,
|
| 1311 |
+
"grad_norm": 8.690220832824707,
|
| 1312 |
+
"learning_rate": 3.033622651049128e-07,
|
| 1313 |
+
"loss": 3.3511,
|
| 1314 |
+
"step": 59500
|
| 1315 |
+
}
|
| 1316 |
+
],
|
| 1317 |
+
"logging_steps": 500,
|
| 1318 |
+
"max_steps": 59835,
|
| 1319 |
+
"num_input_tokens_seen": 0,
|
| 1320 |
+
"num_train_epochs": 1,
|
| 1321 |
+
"save_steps": 2000,
|
| 1322 |
+
"stateful_callbacks": {
|
| 1323 |
+
"TrainerControl": {
|
| 1324 |
+
"args": {
|
| 1325 |
+
"should_epoch_stop": false,
|
| 1326 |
+
"should_evaluate": false,
|
| 1327 |
+
"should_log": false,
|
| 1328 |
+
"should_save": true,
|
| 1329 |
+
"should_training_stop": true
|
| 1330 |
+
},
|
| 1331 |
+
"attributes": {}
|
| 1332 |
+
}
|
| 1333 |
+
},
|
| 1334 |
+
"total_flos": 2.50149494587392e+17,
|
| 1335 |
+
"train_batch_size": 8,
|
| 1336 |
+
"trial_name": null,
|
| 1337 |
+
"trial_params": null
|
| 1338 |
+
}
|
full-finetuning/german-gpt2/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2630695e31d01dc3422c408e2b55254fe3e00fac7a8d16574c59ef8d085fdefc
|
| 3 |
+
size 5304
|
instruct-finetuning/base/LLaMmlein_120M/README.md
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: LSX-UniWue/LLaMmlein_120M
|
| 3 |
+
library_name: peft
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# Model Card for Model ID
|
| 7 |
+
|
| 8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
### Model Description
|
| 15 |
+
|
| 16 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [More Information Needed]
|
| 21 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 22 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 23 |
+
- **Model type:** [More Information Needed]
|
| 24 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 25 |
+
- **License:** [More Information Needed]
|
| 26 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 27 |
+
|
| 28 |
+
### Model Sources [optional]
|
| 29 |
+
|
| 30 |
+
<!-- Provide the basic links for the model. -->
|
| 31 |
+
|
| 32 |
+
- **Repository:** [More Information Needed]
|
| 33 |
+
- **Paper [optional]:** [More Information Needed]
|
| 34 |
+
- **Demo [optional]:** [More Information Needed]
|
| 35 |
+
|
| 36 |
+
## Uses
|
| 37 |
+
|
| 38 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 39 |
+
|
| 40 |
+
### Direct Use
|
| 41 |
+
|
| 42 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 43 |
+
|
| 44 |
+
[More Information Needed]
|
| 45 |
+
|
| 46 |
+
### Downstream Use [optional]
|
| 47 |
+
|
| 48 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 49 |
+
|
| 50 |
+
[More Information Needed]
|
| 51 |
+
|
| 52 |
+
### Out-of-Scope Use
|
| 53 |
+
|
| 54 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 55 |
+
|
| 56 |
+
[More Information Needed]
|
| 57 |
+
|
| 58 |
+
## Bias, Risks, and Limitations
|
| 59 |
+
|
| 60 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 61 |
+
|
| 62 |
+
[More Information Needed]
|
| 63 |
+
|
| 64 |
+
### Recommendations
|
| 65 |
+
|
| 66 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 67 |
+
|
| 68 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 69 |
+
|
| 70 |
+
## How to Get Started with the Model
|
| 71 |
+
|
| 72 |
+
Use the code below to get started with the model.
|
| 73 |
+
|
| 74 |
+
[More Information Needed]
|
| 75 |
+
|
| 76 |
+
## Training Details
|
| 77 |
+
|
| 78 |
+
### Training Data
|
| 79 |
+
|
| 80 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 81 |
+
|
| 82 |
+
[More Information Needed]
|
| 83 |
+
|
| 84 |
+
### Training Procedure
|
| 85 |
+
|
| 86 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 87 |
+
|
| 88 |
+
#### Preprocessing [optional]
|
| 89 |
+
|
| 90 |
+
[More Information Needed]
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
#### Training Hyperparameters
|
| 94 |
+
|
| 95 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 96 |
+
|
| 97 |
+
#### Speeds, Sizes, Times [optional]
|
| 98 |
+
|
| 99 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 100 |
+
|
| 101 |
+
[More Information Needed]
|
| 102 |
+
|
| 103 |
+
## Evaluation
|
| 104 |
+
|
| 105 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 106 |
+
|
| 107 |
+
### Testing Data, Factors & Metrics
|
| 108 |
+
|
| 109 |
+
#### Testing Data
|
| 110 |
+
|
| 111 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 112 |
+
|
| 113 |
+
[More Information Needed]
|
| 114 |
+
|
| 115 |
+
#### Factors
|
| 116 |
+
|
| 117 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 118 |
+
|
| 119 |
+
[More Information Needed]
|
| 120 |
+
|
| 121 |
+
#### Metrics
|
| 122 |
+
|
| 123 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 124 |
+
|
| 125 |
+
[More Information Needed]
|
| 126 |
+
|
| 127 |
+
### Results
|
| 128 |
+
|
| 129 |
+
[More Information Needed]
|
| 130 |
+
|
| 131 |
+
#### Summary
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
|
| 135 |
+
## Model Examination [optional]
|
| 136 |
+
|
| 137 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 138 |
+
|
| 139 |
+
[More Information Needed]
|
| 140 |
+
|
| 141 |
+
## Environmental Impact
|
| 142 |
+
|
| 143 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 144 |
+
|
| 145 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 146 |
+
|
| 147 |
+
- **Hardware Type:** [More Information Needed]
|
| 148 |
+
- **Hours used:** [More Information Needed]
|
| 149 |
+
- **Cloud Provider:** [More Information Needed]
|
| 150 |
+
- **Compute Region:** [More Information Needed]
|
| 151 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 152 |
+
|
| 153 |
+
## Technical Specifications [optional]
|
| 154 |
+
|
| 155 |
+
### Model Architecture and Objective
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
### Compute Infrastructure
|
| 160 |
+
|
| 161 |
+
[More Information Needed]
|
| 162 |
+
|
| 163 |
+
#### Hardware
|
| 164 |
+
|
| 165 |
+
[More Information Needed]
|
| 166 |
+
|
| 167 |
+
#### Software
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
## Citation [optional]
|
| 172 |
+
|
| 173 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 174 |
+
|
| 175 |
+
**BibTeX:**
|
| 176 |
+
|
| 177 |
+
[More Information Needed]
|
| 178 |
+
|
| 179 |
+
**APA:**
|
| 180 |
+
|
| 181 |
+
[More Information Needed]
|
| 182 |
+
|
| 183 |
+
## Glossary [optional]
|
| 184 |
+
|
| 185 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 186 |
+
|
| 187 |
+
[More Information Needed]
|
| 188 |
+
|
| 189 |
+
## More Information [optional]
|
| 190 |
+
|
| 191 |
+
[More Information Needed]
|
| 192 |
+
|
| 193 |
+
## Model Card Authors [optional]
|
| 194 |
+
|
| 195 |
+
[More Information Needed]
|
| 196 |
+
|
| 197 |
+
## Model Card Contact
|
| 198 |
+
|
| 199 |
+
[More Information Needed]
|
| 200 |
+
### Framework versions
|
| 201 |
+
|
| 202 |
+
- PEFT 0.13.2
|
instruct-finetuning/base/LLaMmlein_120M/adapter_config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "LSX-UniWue/LLaMmlein_120M",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layer_replication": null,
|
| 10 |
+
"layers_pattern": null,
|
| 11 |
+
"layers_to_transform": null,
|
| 12 |
+
"loftq_config": {},
|
| 13 |
+
"lora_alpha": 4,
|
| 14 |
+
"lora_dropout": 0.1,
|
| 15 |
+
"megatron_config": null,
|
| 16 |
+
"megatron_core": "megatron.core",
|
| 17 |
+
"modules_to_save": null,
|
| 18 |
+
"peft_type": "LORA",
|
| 19 |
+
"r": 32,
|
| 20 |
+
"rank_pattern": {},
|
| 21 |
+
"revision": null,
|
| 22 |
+
"target_modules": [
|
| 23 |
+
"up_proj",
|
| 24 |
+
"gate_proj",
|
| 25 |
+
"o_proj",
|
| 26 |
+
"v_proj",
|
| 27 |
+
"q_proj",
|
| 28 |
+
"k_proj",
|
| 29 |
+
"lm_head",
|
| 30 |
+
"down_proj"
|
| 31 |
+
],
|
| 32 |
+
"task_type": "CAUSAL_LM",
|
| 33 |
+
"use_dora": false,
|
| 34 |
+
"use_rslora": false
|
| 35 |
+
}
|
instruct-finetuning/base/LLaMmlein_120M/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a839a6e5514599a2a43cc1f29f435b0f9a726019a5cdec1abd0a91a44ef66eb8
|
| 3 |
+
size 123361608
|
instruct-finetuning/base/LLaMmlein_120M/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:82f506a64c04df6456eb1373f4f11de4e0552f2cd77656720971d58b429541a8
|
| 3 |
+
size 850871994
|
instruct-finetuning/base/LLaMmlein_120M/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e42ccd97e89dcb0ecb4e53ebfd9919fde6e4d4c5c0a12a689387afb57967751d
|
| 3 |
+
size 14244
|
instruct-finetuning/base/LLaMmlein_120M/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0af176d761d71fce3fbce7001f4850782b022af8f40338e8e88b22363a32018f
|
| 3 |
+
size 1064
|
instruct-finetuning/base/LLaMmlein_120M/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
instruct-finetuning/base/LLaMmlein_120M/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
instruct-finetuning/base/LLaMmlein_120M/tokenizer_config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"additional_special_tokens": [],
|
| 32 |
+
"bos_token": "<s>",
|
| 33 |
+
"clean_up_tokenization_spaces": false,
|
| 34 |
+
"eos_token": "</s>",
|
| 35 |
+
"legacy": false,
|
| 36 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 37 |
+
"pad_token": "</s>",
|
| 38 |
+
"padding_side": "left",
|
| 39 |
+
"sp_model_kwargs": {},
|
| 40 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 41 |
+
"unk_token": "<unk>",
|
| 42 |
+
"use_default_system_prompt": true
|
| 43 |
+
}
|
instruct-finetuning/base/LLaMmlein_120M/trainer_state.json
ADDED
|
@@ -0,0 +1,1593 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 6.343393325805664,
|
| 3 |
+
"best_model_checkpoint": "./models/instruct-finetuning/base/LLaMmlein_120M/checkpoint-100",
|
| 4 |
+
"epoch": 2000.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 2000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 10.0,
|
| 13 |
+
"grad_norm": 7.257441997528076,
|
| 14 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 15 |
+
"loss": 4.0845,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 20.0,
|
| 20 |
+
"grad_norm": 6.133562088012695,
|
| 21 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 22 |
+
"loss": 1.8532,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 30.0,
|
| 27 |
+
"grad_norm": 2.961812734603882,
|
| 28 |
+
"learning_rate": 8.999999999999999e-05,
|
| 29 |
+
"loss": 0.3387,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 40.0,
|
| 34 |
+
"grad_norm": 1.0544193983078003,
|
| 35 |
+
"learning_rate": 0.00011999999999999999,
|
| 36 |
+
"loss": 0.0728,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 50.0,
|
| 41 |
+
"grad_norm": 0.11237581074237823,
|
| 42 |
+
"learning_rate": 0.00015,
|
| 43 |
+
"loss": 0.0122,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 60.0,
|
| 48 |
+
"grad_norm": 0.04610421508550644,
|
| 49 |
+
"learning_rate": 0.00017999999999999998,
|
| 50 |
+
"loss": 0.0066,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 70.0,
|
| 55 |
+
"grad_norm": 0.031031867489218712,
|
| 56 |
+
"learning_rate": 0.00020999999999999998,
|
| 57 |
+
"loss": 0.0057,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 80.0,
|
| 62 |
+
"grad_norm": 0.017072567716240883,
|
| 63 |
+
"learning_rate": 0.00023999999999999998,
|
| 64 |
+
"loss": 0.0055,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 90.0,
|
| 69 |
+
"grad_norm": 0.00996345654129982,
|
| 70 |
+
"learning_rate": 0.00027,
|
| 71 |
+
"loss": 0.0055,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 100.0,
|
| 76 |
+
"grad_norm": 0.009029845707118511,
|
| 77 |
+
"learning_rate": 0.0003,
|
| 78 |
+
"loss": 0.0055,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 100.0,
|
| 83 |
+
"eval_loss": 6.343393325805664,
|
| 84 |
+
"eval_runtime": 0.0563,
|
| 85 |
+
"eval_samples_per_second": 283.996,
|
| 86 |
+
"eval_steps_per_second": 17.75,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 110.0,
|
| 91 |
+
"grad_norm": 0.0047378516755998135,
|
| 92 |
+
"learning_rate": 0.00029842105263157894,
|
| 93 |
+
"loss": 0.0055,
|
| 94 |
+
"step": 110
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 120.0,
|
| 98 |
+
"grad_norm": 0.003162564244121313,
|
| 99 |
+
"learning_rate": 0.00029684210526315785,
|
| 100 |
+
"loss": 0.0055,
|
| 101 |
+
"step": 120
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 130.0,
|
| 105 |
+
"grad_norm": 0.009896782226860523,
|
| 106 |
+
"learning_rate": 0.0002952631578947368,
|
| 107 |
+
"loss": 0.0054,
|
| 108 |
+
"step": 130
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 140.0,
|
| 112 |
+
"grad_norm": 0.005971991922706366,
|
| 113 |
+
"learning_rate": 0.0002936842105263158,
|
| 114 |
+
"loss": 0.0055,
|
| 115 |
+
"step": 140
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 150.0,
|
| 119 |
+
"grad_norm": 0.006593422032892704,
|
| 120 |
+
"learning_rate": 0.0002921052631578947,
|
| 121 |
+
"loss": 0.0054,
|
| 122 |
+
"step": 150
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 160.0,
|
| 126 |
+
"grad_norm": 0.01387177873402834,
|
| 127 |
+
"learning_rate": 0.00029052631578947366,
|
| 128 |
+
"loss": 0.0054,
|
| 129 |
+
"step": 160
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 170.0,
|
| 133 |
+
"grad_norm": 0.01074151135981083,
|
| 134 |
+
"learning_rate": 0.00028894736842105263,
|
| 135 |
+
"loss": 0.0055,
|
| 136 |
+
"step": 170
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 180.0,
|
| 140 |
+
"grad_norm": 0.004622858017683029,
|
| 141 |
+
"learning_rate": 0.00028736842105263154,
|
| 142 |
+
"loss": 0.0054,
|
| 143 |
+
"step": 180
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 190.0,
|
| 147 |
+
"grad_norm": 0.0028191779274493456,
|
| 148 |
+
"learning_rate": 0.0002857894736842105,
|
| 149 |
+
"loss": 0.0054,
|
| 150 |
+
"step": 190
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 200.0,
|
| 154 |
+
"grad_norm": 0.007395514752715826,
|
| 155 |
+
"learning_rate": 0.0002842105263157894,
|
| 156 |
+
"loss": 0.0054,
|
| 157 |
+
"step": 200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 200.0,
|
| 161 |
+
"eval_loss": 6.482385158538818,
|
| 162 |
+
"eval_runtime": 0.0567,
|
| 163 |
+
"eval_samples_per_second": 282.327,
|
| 164 |
+
"eval_steps_per_second": 17.645,
|
| 165 |
+
"step": 200
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 210.0,
|
| 169 |
+
"grad_norm": 0.002630041679367423,
|
| 170 |
+
"learning_rate": 0.0002826315789473684,
|
| 171 |
+
"loss": 0.0054,
|
| 172 |
+
"step": 210
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 220.0,
|
| 176 |
+
"grad_norm": 0.009388357400894165,
|
| 177 |
+
"learning_rate": 0.00028105263157894735,
|
| 178 |
+
"loss": 0.0054,
|
| 179 |
+
"step": 220
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 230.0,
|
| 183 |
+
"grad_norm": 0.012319635599851608,
|
| 184 |
+
"learning_rate": 0.0002794736842105263,
|
| 185 |
+
"loss": 0.0054,
|
| 186 |
+
"step": 230
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 240.0,
|
| 190 |
+
"grad_norm": 0.008075610734522343,
|
| 191 |
+
"learning_rate": 0.00027789473684210523,
|
| 192 |
+
"loss": 0.0054,
|
| 193 |
+
"step": 240
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 250.0,
|
| 197 |
+
"grad_norm": 0.0070420210249722,
|
| 198 |
+
"learning_rate": 0.0002763157894736842,
|
| 199 |
+
"loss": 0.0054,
|
| 200 |
+
"step": 250
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 260.0,
|
| 204 |
+
"grad_norm": 0.0077827088534832,
|
| 205 |
+
"learning_rate": 0.0002747368421052631,
|
| 206 |
+
"loss": 0.0054,
|
| 207 |
+
"step": 260
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 270.0,
|
| 211 |
+
"grad_norm": 0.009190627373754978,
|
| 212 |
+
"learning_rate": 0.00027315789473684207,
|
| 213 |
+
"loss": 0.0054,
|
| 214 |
+
"step": 270
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 280.0,
|
| 218 |
+
"grad_norm": 0.006282865069806576,
|
| 219 |
+
"learning_rate": 0.00027157894736842104,
|
| 220 |
+
"loss": 0.0054,
|
| 221 |
+
"step": 280
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 290.0,
|
| 225 |
+
"grad_norm": 0.00803539901971817,
|
| 226 |
+
"learning_rate": 0.00027,
|
| 227 |
+
"loss": 0.0054,
|
| 228 |
+
"step": 290
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 300.0,
|
| 232 |
+
"grad_norm": 0.005915256217122078,
|
| 233 |
+
"learning_rate": 0.0002684210526315789,
|
| 234 |
+
"loss": 0.0054,
|
| 235 |
+
"step": 300
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 300.0,
|
| 239 |
+
"eval_loss": 6.544665336608887,
|
| 240 |
+
"eval_runtime": 0.0575,
|
| 241 |
+
"eval_samples_per_second": 278.46,
|
| 242 |
+
"eval_steps_per_second": 17.404,
|
| 243 |
+
"step": 300
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 310.0,
|
| 247 |
+
"grad_norm": 0.00953945517539978,
|
| 248 |
+
"learning_rate": 0.0002668421052631579,
|
| 249 |
+
"loss": 0.0054,
|
| 250 |
+
"step": 310
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 320.0,
|
| 254 |
+
"grad_norm": 0.0184449665248394,
|
| 255 |
+
"learning_rate": 0.0002652631578947368,
|
| 256 |
+
"loss": 0.0054,
|
| 257 |
+
"step": 320
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 330.0,
|
| 261 |
+
"grad_norm": 0.009298360906541348,
|
| 262 |
+
"learning_rate": 0.00026368421052631576,
|
| 263 |
+
"loss": 0.0054,
|
| 264 |
+
"step": 330
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 340.0,
|
| 268 |
+
"grad_norm": 0.007060893811285496,
|
| 269 |
+
"learning_rate": 0.0002621052631578947,
|
| 270 |
+
"loss": 0.0054,
|
| 271 |
+
"step": 340
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 350.0,
|
| 275 |
+
"grad_norm": 0.005190024618059397,
|
| 276 |
+
"learning_rate": 0.0002605263157894737,
|
| 277 |
+
"loss": 0.0054,
|
| 278 |
+
"step": 350
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 360.0,
|
| 282 |
+
"grad_norm": 0.007410724181681871,
|
| 283 |
+
"learning_rate": 0.0002589473684210526,
|
| 284 |
+
"loss": 0.0054,
|
| 285 |
+
"step": 360
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 370.0,
|
| 289 |
+
"grad_norm": 0.0066852509044110775,
|
| 290 |
+
"learning_rate": 0.00025736842105263157,
|
| 291 |
+
"loss": 0.0054,
|
| 292 |
+
"step": 370
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 380.0,
|
| 296 |
+
"grad_norm": 0.00933538842946291,
|
| 297 |
+
"learning_rate": 0.0002557894736842105,
|
| 298 |
+
"loss": 0.0054,
|
| 299 |
+
"step": 380
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 390.0,
|
| 303 |
+
"grad_norm": 0.006421348080039024,
|
| 304 |
+
"learning_rate": 0.00025421052631578945,
|
| 305 |
+
"loss": 0.0054,
|
| 306 |
+
"step": 390
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 400.0,
|
| 310 |
+
"grad_norm": 0.013186627998948097,
|
| 311 |
+
"learning_rate": 0.00025263157894736836,
|
| 312 |
+
"loss": 0.0054,
|
| 313 |
+
"step": 400
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 400.0,
|
| 317 |
+
"eval_loss": 6.592121601104736,
|
| 318 |
+
"eval_runtime": 0.0566,
|
| 319 |
+
"eval_samples_per_second": 282.605,
|
| 320 |
+
"eval_steps_per_second": 17.663,
|
| 321 |
+
"step": 400
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 410.0,
|
| 325 |
+
"grad_norm": 0.008023693226277828,
|
| 326 |
+
"learning_rate": 0.0002510526315789474,
|
| 327 |
+
"loss": 0.0054,
|
| 328 |
+
"step": 410
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 420.0,
|
| 332 |
+
"grad_norm": 0.006304403301328421,
|
| 333 |
+
"learning_rate": 0.0002494736842105263,
|
| 334 |
+
"loss": 0.0054,
|
| 335 |
+
"step": 420
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 430.0,
|
| 339 |
+
"grad_norm": 0.006279406137764454,
|
| 340 |
+
"learning_rate": 0.00024789473684210526,
|
| 341 |
+
"loss": 0.0054,
|
| 342 |
+
"step": 430
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 440.0,
|
| 346 |
+
"grad_norm": 0.005663767457008362,
|
| 347 |
+
"learning_rate": 0.00024631578947368417,
|
| 348 |
+
"loss": 0.0054,
|
| 349 |
+
"step": 440
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 450.0,
|
| 353 |
+
"grad_norm": 0.007969029247760773,
|
| 354 |
+
"learning_rate": 0.00024473684210526314,
|
| 355 |
+
"loss": 0.0054,
|
| 356 |
+
"step": 450
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 460.0,
|
| 360 |
+
"grad_norm": 0.004170434549450874,
|
| 361 |
+
"learning_rate": 0.00024315789473684207,
|
| 362 |
+
"loss": 0.0054,
|
| 363 |
+
"step": 460
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 470.0,
|
| 367 |
+
"grad_norm": 0.007959424518048763,
|
| 368 |
+
"learning_rate": 0.000241578947368421,
|
| 369 |
+
"loss": 0.0054,
|
| 370 |
+
"step": 470
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 480.0,
|
| 374 |
+
"grad_norm": 0.007390080485492945,
|
| 375 |
+
"learning_rate": 0.00023999999999999998,
|
| 376 |
+
"loss": 0.0054,
|
| 377 |
+
"step": 480
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 490.0,
|
| 381 |
+
"grad_norm": 0.006555990315973759,
|
| 382 |
+
"learning_rate": 0.00023842105263157895,
|
| 383 |
+
"loss": 0.0054,
|
| 384 |
+
"step": 490
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 500.0,
|
| 388 |
+
"grad_norm": 0.007358189672231674,
|
| 389 |
+
"learning_rate": 0.00023684210526315788,
|
| 390 |
+
"loss": 0.0054,
|
| 391 |
+
"step": 500
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 500.0,
|
| 395 |
+
"eval_loss": 6.633509159088135,
|
| 396 |
+
"eval_runtime": 0.0562,
|
| 397 |
+
"eval_samples_per_second": 284.773,
|
| 398 |
+
"eval_steps_per_second": 17.798,
|
| 399 |
+
"step": 500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 510.0,
|
| 403 |
+
"grad_norm": 0.008251392282545567,
|
| 404 |
+
"learning_rate": 0.00023526315789473682,
|
| 405 |
+
"loss": 0.0054,
|
| 406 |
+
"step": 510
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 520.0,
|
| 410 |
+
"grad_norm": 0.0032486410345882177,
|
| 411 |
+
"learning_rate": 0.00023368421052631576,
|
| 412 |
+
"loss": 0.0054,
|
| 413 |
+
"step": 520
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 530.0,
|
| 417 |
+
"grad_norm": 0.01079108752310276,
|
| 418 |
+
"learning_rate": 0.0002321052631578947,
|
| 419 |
+
"loss": 0.0054,
|
| 420 |
+
"step": 530
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 540.0,
|
| 424 |
+
"grad_norm": 0.006488645449280739,
|
| 425 |
+
"learning_rate": 0.00023052631578947364,
|
| 426 |
+
"loss": 0.0054,
|
| 427 |
+
"step": 540
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 550.0,
|
| 431 |
+
"grad_norm": 0.006889748852699995,
|
| 432 |
+
"learning_rate": 0.00022894736842105263,
|
| 433 |
+
"loss": 0.0054,
|
| 434 |
+
"step": 550
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 560.0,
|
| 438 |
+
"grad_norm": 0.005966802127659321,
|
| 439 |
+
"learning_rate": 0.00022736842105263157,
|
| 440 |
+
"loss": 0.0054,
|
| 441 |
+
"step": 560
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 570.0,
|
| 445 |
+
"grad_norm": 0.00871422328054905,
|
| 446 |
+
"learning_rate": 0.0002257894736842105,
|
| 447 |
+
"loss": 0.0054,
|
| 448 |
+
"step": 570
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 580.0,
|
| 452 |
+
"grad_norm": 0.007486347574740648,
|
| 453 |
+
"learning_rate": 0.00022421052631578945,
|
| 454 |
+
"loss": 0.0054,
|
| 455 |
+
"step": 580
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 590.0,
|
| 459 |
+
"grad_norm": 0.006703991908580065,
|
| 460 |
+
"learning_rate": 0.0002226315789473684,
|
| 461 |
+
"loss": 0.0054,
|
| 462 |
+
"step": 590
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 600.0,
|
| 466 |
+
"grad_norm": 0.0069758091121912,
|
| 467 |
+
"learning_rate": 0.00022105263157894733,
|
| 468 |
+
"loss": 0.0054,
|
| 469 |
+
"step": 600
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 600.0,
|
| 473 |
+
"eval_loss": 6.659405708312988,
|
| 474 |
+
"eval_runtime": 0.0561,
|
| 475 |
+
"eval_samples_per_second": 284.975,
|
| 476 |
+
"eval_steps_per_second": 17.811,
|
| 477 |
+
"step": 600
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 610.0,
|
| 481 |
+
"grad_norm": 0.006505531724542379,
|
| 482 |
+
"learning_rate": 0.00021947368421052632,
|
| 483 |
+
"loss": 0.0054,
|
| 484 |
+
"step": 610
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 620.0,
|
| 488 |
+
"grad_norm": 0.006151702255010605,
|
| 489 |
+
"learning_rate": 0.00021789473684210526,
|
| 490 |
+
"loss": 0.0054,
|
| 491 |
+
"step": 620
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 630.0,
|
| 495 |
+
"grad_norm": 0.006619542371481657,
|
| 496 |
+
"learning_rate": 0.0002163157894736842,
|
| 497 |
+
"loss": 0.0054,
|
| 498 |
+
"step": 630
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 640.0,
|
| 502 |
+
"grad_norm": 0.0047877514734864235,
|
| 503 |
+
"learning_rate": 0.00021473684210526314,
|
| 504 |
+
"loss": 0.0054,
|
| 505 |
+
"step": 640
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 650.0,
|
| 509 |
+
"grad_norm": 0.005860659293830395,
|
| 510 |
+
"learning_rate": 0.00021315789473684208,
|
| 511 |
+
"loss": 0.0054,
|
| 512 |
+
"step": 650
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 660.0,
|
| 516 |
+
"grad_norm": 0.0046273041516542435,
|
| 517 |
+
"learning_rate": 0.00021157894736842102,
|
| 518 |
+
"loss": 0.0054,
|
| 519 |
+
"step": 660
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 670.0,
|
| 523 |
+
"grad_norm": 0.00666681258007884,
|
| 524 |
+
"learning_rate": 0.00020999999999999998,
|
| 525 |
+
"loss": 0.0054,
|
| 526 |
+
"step": 670
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 680.0,
|
| 530 |
+
"grad_norm": 0.0070931087248027325,
|
| 531 |
+
"learning_rate": 0.00020842105263157895,
|
| 532 |
+
"loss": 0.0054,
|
| 533 |
+
"step": 680
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 690.0,
|
| 537 |
+
"grad_norm": 0.005947079975157976,
|
| 538 |
+
"learning_rate": 0.0002068421052631579,
|
| 539 |
+
"loss": 0.0054,
|
| 540 |
+
"step": 690
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 700.0,
|
| 544 |
+
"grad_norm": 0.007008504122495651,
|
| 545 |
+
"learning_rate": 0.00020526315789473683,
|
| 546 |
+
"loss": 0.0054,
|
| 547 |
+
"step": 700
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 700.0,
|
| 551 |
+
"eval_loss": 6.695831775665283,
|
| 552 |
+
"eval_runtime": 0.0572,
|
| 553 |
+
"eval_samples_per_second": 279.963,
|
| 554 |
+
"eval_steps_per_second": 17.498,
|
| 555 |
+
"step": 700
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 710.0,
|
| 559 |
+
"grad_norm": 0.00678026070818305,
|
| 560 |
+
"learning_rate": 0.00020368421052631576,
|
| 561 |
+
"loss": 0.0054,
|
| 562 |
+
"step": 710
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 720.0,
|
| 566 |
+
"grad_norm": 0.005221761297434568,
|
| 567 |
+
"learning_rate": 0.0002021052631578947,
|
| 568 |
+
"loss": 0.0054,
|
| 569 |
+
"step": 720
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 730.0,
|
| 573 |
+
"grad_norm": 0.005231580231338739,
|
| 574 |
+
"learning_rate": 0.00020052631578947367,
|
| 575 |
+
"loss": 0.0054,
|
| 576 |
+
"step": 730
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 740.0,
|
| 580 |
+
"grad_norm": 0.006937530357390642,
|
| 581 |
+
"learning_rate": 0.0001989473684210526,
|
| 582 |
+
"loss": 0.0054,
|
| 583 |
+
"step": 740
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 750.0,
|
| 587 |
+
"grad_norm": 0.007865114137530327,
|
| 588 |
+
"learning_rate": 0.00019736842105263157,
|
| 589 |
+
"loss": 0.0054,
|
| 590 |
+
"step": 750
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 760.0,
|
| 594 |
+
"grad_norm": 0.014306128025054932,
|
| 595 |
+
"learning_rate": 0.0001957894736842105,
|
| 596 |
+
"loss": 0.0054,
|
| 597 |
+
"step": 760
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 770.0,
|
| 601 |
+
"grad_norm": 0.006088100839406252,
|
| 602 |
+
"learning_rate": 0.00019421052631578945,
|
| 603 |
+
"loss": 0.0054,
|
| 604 |
+
"step": 770
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 780.0,
|
| 608 |
+
"grad_norm": 0.004863563925027847,
|
| 609 |
+
"learning_rate": 0.0001926315789473684,
|
| 610 |
+
"loss": 0.0054,
|
| 611 |
+
"step": 780
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 790.0,
|
| 615 |
+
"grad_norm": 0.005216915160417557,
|
| 616 |
+
"learning_rate": 0.00019105263157894736,
|
| 617 |
+
"loss": 0.0054,
|
| 618 |
+
"step": 790
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 800.0,
|
| 622 |
+
"grad_norm": 0.0025104102678596973,
|
| 623 |
+
"learning_rate": 0.0001894736842105263,
|
| 624 |
+
"loss": 0.0054,
|
| 625 |
+
"step": 800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 800.0,
|
| 629 |
+
"eval_loss": 6.713556289672852,
|
| 630 |
+
"eval_runtime": 0.0569,
|
| 631 |
+
"eval_samples_per_second": 281.395,
|
| 632 |
+
"eval_steps_per_second": 17.587,
|
| 633 |
+
"step": 800
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 810.0,
|
| 637 |
+
"grad_norm": 0.004186335019767284,
|
| 638 |
+
"learning_rate": 0.00018789473684210524,
|
| 639 |
+
"loss": 0.0054,
|
| 640 |
+
"step": 810
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 820.0,
|
| 644 |
+
"grad_norm": 0.005255494732409716,
|
| 645 |
+
"learning_rate": 0.0001863157894736842,
|
| 646 |
+
"loss": 0.0054,
|
| 647 |
+
"step": 820
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 830.0,
|
| 651 |
+
"grad_norm": 0.005517592187970877,
|
| 652 |
+
"learning_rate": 0.00018473684210526314,
|
| 653 |
+
"loss": 0.0054,
|
| 654 |
+
"step": 830
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 840.0,
|
| 658 |
+
"grad_norm": 0.0052512213587760925,
|
| 659 |
+
"learning_rate": 0.00018315789473684208,
|
| 660 |
+
"loss": 0.0054,
|
| 661 |
+
"step": 840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 850.0,
|
| 665 |
+
"grad_norm": 0.005355685483664274,
|
| 666 |
+
"learning_rate": 0.00018157894736842105,
|
| 667 |
+
"loss": 0.0054,
|
| 668 |
+
"step": 850
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 860.0,
|
| 672 |
+
"grad_norm": 0.0037962698843330145,
|
| 673 |
+
"learning_rate": 0.00017999999999999998,
|
| 674 |
+
"loss": 0.0054,
|
| 675 |
+
"step": 860
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 870.0,
|
| 679 |
+
"grad_norm": 0.0035633454099297523,
|
| 680 |
+
"learning_rate": 0.00017842105263157892,
|
| 681 |
+
"loss": 0.0054,
|
| 682 |
+
"step": 870
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 880.0,
|
| 686 |
+
"grad_norm": 0.00413065729662776,
|
| 687 |
+
"learning_rate": 0.00017684210526315786,
|
| 688 |
+
"loss": 0.0054,
|
| 689 |
+
"step": 880
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 890.0,
|
| 693 |
+
"grad_norm": 0.0030881876591593027,
|
| 694 |
+
"learning_rate": 0.00017526315789473683,
|
| 695 |
+
"loss": 0.0054,
|
| 696 |
+
"step": 890
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 900.0,
|
| 700 |
+
"grad_norm": 0.005164352711290121,
|
| 701 |
+
"learning_rate": 0.0001736842105263158,
|
| 702 |
+
"loss": 0.0054,
|
| 703 |
+
"step": 900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 900.0,
|
| 707 |
+
"eval_loss": 6.732208251953125,
|
| 708 |
+
"eval_runtime": 0.0561,
|
| 709 |
+
"eval_samples_per_second": 285.23,
|
| 710 |
+
"eval_steps_per_second": 17.827,
|
| 711 |
+
"step": 900
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 910.0,
|
| 715 |
+
"grad_norm": 0.0037032829131931067,
|
| 716 |
+
"learning_rate": 0.00017210526315789473,
|
| 717 |
+
"loss": 0.0054,
|
| 718 |
+
"step": 910
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 920.0,
|
| 722 |
+
"grad_norm": 0.009146653115749359,
|
| 723 |
+
"learning_rate": 0.00017052631578947367,
|
| 724 |
+
"loss": 0.0054,
|
| 725 |
+
"step": 920
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 930.0,
|
| 729 |
+
"grad_norm": 0.004405967425554991,
|
| 730 |
+
"learning_rate": 0.0001689473684210526,
|
| 731 |
+
"loss": 0.0054,
|
| 732 |
+
"step": 930
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 940.0,
|
| 736 |
+
"grad_norm": 0.004398548975586891,
|
| 737 |
+
"learning_rate": 0.00016736842105263155,
|
| 738 |
+
"loss": 0.0054,
|
| 739 |
+
"step": 940
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 950.0,
|
| 743 |
+
"grad_norm": 0.004243563394993544,
|
| 744 |
+
"learning_rate": 0.00016578947368421052,
|
| 745 |
+
"loss": 0.0054,
|
| 746 |
+
"step": 950
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 960.0,
|
| 750 |
+
"grad_norm": 0.004099991172552109,
|
| 751 |
+
"learning_rate": 0.00016421052631578948,
|
| 752 |
+
"loss": 0.0054,
|
| 753 |
+
"step": 960
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 970.0,
|
| 757 |
+
"grad_norm": 0.004985535517334938,
|
| 758 |
+
"learning_rate": 0.00016263157894736842,
|
| 759 |
+
"loss": 0.0054,
|
| 760 |
+
"step": 970
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 980.0,
|
| 764 |
+
"grad_norm": 0.005821101367473602,
|
| 765 |
+
"learning_rate": 0.00016105263157894736,
|
| 766 |
+
"loss": 0.0054,
|
| 767 |
+
"step": 980
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 990.0,
|
| 771 |
+
"grad_norm": 0.006319540087133646,
|
| 772 |
+
"learning_rate": 0.0001594736842105263,
|
| 773 |
+
"loss": 0.0054,
|
| 774 |
+
"step": 990
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1000.0,
|
| 778 |
+
"grad_norm": 0.004705901723355055,
|
| 779 |
+
"learning_rate": 0.00015789473684210524,
|
| 780 |
+
"loss": 0.0054,
|
| 781 |
+
"step": 1000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1000.0,
|
| 785 |
+
"eval_loss": 6.75187349319458,
|
| 786 |
+
"eval_runtime": 0.0567,
|
| 787 |
+
"eval_samples_per_second": 282.37,
|
| 788 |
+
"eval_steps_per_second": 17.648,
|
| 789 |
+
"step": 1000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 1010.0,
|
| 793 |
+
"grad_norm": 0.0037577070761471987,
|
| 794 |
+
"learning_rate": 0.00015631578947368418,
|
| 795 |
+
"loss": 0.0054,
|
| 796 |
+
"step": 1010
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 1020.0,
|
| 800 |
+
"grad_norm": 0.0032031466253101826,
|
| 801 |
+
"learning_rate": 0.00015473684210526317,
|
| 802 |
+
"loss": 0.0054,
|
| 803 |
+
"step": 1020
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1030.0,
|
| 807 |
+
"grad_norm": 0.0021940330043435097,
|
| 808 |
+
"learning_rate": 0.0001531578947368421,
|
| 809 |
+
"loss": 0.0054,
|
| 810 |
+
"step": 1030
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1040.0,
|
| 814 |
+
"grad_norm": 0.0029385159723460674,
|
| 815 |
+
"learning_rate": 0.00015157894736842105,
|
| 816 |
+
"loss": 0.0054,
|
| 817 |
+
"step": 1040
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 1050.0,
|
| 821 |
+
"grad_norm": 0.003867937019094825,
|
| 822 |
+
"learning_rate": 0.00015,
|
| 823 |
+
"loss": 0.0054,
|
| 824 |
+
"step": 1050
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1060.0,
|
| 828 |
+
"grad_norm": 0.003548321081325412,
|
| 829 |
+
"learning_rate": 0.00014842105263157893,
|
| 830 |
+
"loss": 0.0054,
|
| 831 |
+
"step": 1060
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 1070.0,
|
| 835 |
+
"grad_norm": 0.004459850490093231,
|
| 836 |
+
"learning_rate": 0.0001468421052631579,
|
| 837 |
+
"loss": 0.0054,
|
| 838 |
+
"step": 1070
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1080.0,
|
| 842 |
+
"grad_norm": 0.008197680115699768,
|
| 843 |
+
"learning_rate": 0.00014526315789473683,
|
| 844 |
+
"loss": 0.0054,
|
| 845 |
+
"step": 1080
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1090.0,
|
| 849 |
+
"grad_norm": 0.0062789651565253735,
|
| 850 |
+
"learning_rate": 0.00014368421052631577,
|
| 851 |
+
"loss": 0.0054,
|
| 852 |
+
"step": 1090
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 1100.0,
|
| 856 |
+
"grad_norm": 0.003925560973584652,
|
| 857 |
+
"learning_rate": 0.0001421052631578947,
|
| 858 |
+
"loss": 0.0054,
|
| 859 |
+
"step": 1100
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 1100.0,
|
| 863 |
+
"eval_loss": 6.772552013397217,
|
| 864 |
+
"eval_runtime": 0.0757,
|
| 865 |
+
"eval_samples_per_second": 211.273,
|
| 866 |
+
"eval_steps_per_second": 13.205,
|
| 867 |
+
"step": 1100
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 1110.0,
|
| 871 |
+
"grad_norm": 0.003982131835073233,
|
| 872 |
+
"learning_rate": 0.00014052631578947367,
|
| 873 |
+
"loss": 0.0054,
|
| 874 |
+
"step": 1110
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 1120.0,
|
| 878 |
+
"grad_norm": 0.003327371319755912,
|
| 879 |
+
"learning_rate": 0.00013894736842105261,
|
| 880 |
+
"loss": 0.0054,
|
| 881 |
+
"step": 1120
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1130.0,
|
| 885 |
+
"grad_norm": 0.0032421585638076067,
|
| 886 |
+
"learning_rate": 0.00013736842105263155,
|
| 887 |
+
"loss": 0.0054,
|
| 888 |
+
"step": 1130
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 1140.0,
|
| 892 |
+
"grad_norm": 0.004894908983260393,
|
| 893 |
+
"learning_rate": 0.00013578947368421052,
|
| 894 |
+
"loss": 0.0054,
|
| 895 |
+
"step": 1140
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 1150.0,
|
| 899 |
+
"grad_norm": 0.0032253412064164877,
|
| 900 |
+
"learning_rate": 0.00013421052631578946,
|
| 901 |
+
"loss": 0.0054,
|
| 902 |
+
"step": 1150
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1160.0,
|
| 906 |
+
"grad_norm": 0.0022986368276178837,
|
| 907 |
+
"learning_rate": 0.0001326315789473684,
|
| 908 |
+
"loss": 0.0054,
|
| 909 |
+
"step": 1160
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 1170.0,
|
| 913 |
+
"grad_norm": 0.004993563052266836,
|
| 914 |
+
"learning_rate": 0.00013105263157894736,
|
| 915 |
+
"loss": 0.0054,
|
| 916 |
+
"step": 1170
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 1180.0,
|
| 920 |
+
"grad_norm": 0.0036353906616568565,
|
| 921 |
+
"learning_rate": 0.0001294736842105263,
|
| 922 |
+
"loss": 0.0054,
|
| 923 |
+
"step": 1180
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1190.0,
|
| 927 |
+
"grad_norm": 0.00476466491818428,
|
| 928 |
+
"learning_rate": 0.00012789473684210524,
|
| 929 |
+
"loss": 0.0054,
|
| 930 |
+
"step": 1190
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 1200.0,
|
| 934 |
+
"grad_norm": 0.003141681896522641,
|
| 935 |
+
"learning_rate": 0.00012631578947368418,
|
| 936 |
+
"loss": 0.0054,
|
| 937 |
+
"step": 1200
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 1200.0,
|
| 941 |
+
"eval_loss": 6.784540176391602,
|
| 942 |
+
"eval_runtime": 0.0567,
|
| 943 |
+
"eval_samples_per_second": 282.119,
|
| 944 |
+
"eval_steps_per_second": 17.632,
|
| 945 |
+
"step": 1200
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 1210.0,
|
| 949 |
+
"grad_norm": 0.004309048876166344,
|
| 950 |
+
"learning_rate": 0.00012473684210526315,
|
| 951 |
+
"loss": 0.0054,
|
| 952 |
+
"step": 1210
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 1220.0,
|
| 956 |
+
"grad_norm": 0.004496368113905191,
|
| 957 |
+
"learning_rate": 0.00012315789473684208,
|
| 958 |
+
"loss": 0.0054,
|
| 959 |
+
"step": 1220
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 1230.0,
|
| 963 |
+
"grad_norm": 0.0035161725245416164,
|
| 964 |
+
"learning_rate": 0.00012157894736842104,
|
| 965 |
+
"loss": 0.0054,
|
| 966 |
+
"step": 1230
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 1240.0,
|
| 970 |
+
"grad_norm": 0.002308123977854848,
|
| 971 |
+
"learning_rate": 0.00011999999999999999,
|
| 972 |
+
"loss": 0.0054,
|
| 973 |
+
"step": 1240
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 1250.0,
|
| 977 |
+
"grad_norm": 0.003306175349280238,
|
| 978 |
+
"learning_rate": 0.00011842105263157894,
|
| 979 |
+
"loss": 0.0054,
|
| 980 |
+
"step": 1250
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1260.0,
|
| 984 |
+
"grad_norm": 0.0042304969392716885,
|
| 985 |
+
"learning_rate": 0.00011684210526315788,
|
| 986 |
+
"loss": 0.0054,
|
| 987 |
+
"step": 1260
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 1270.0,
|
| 991 |
+
"grad_norm": 0.0038438441697508097,
|
| 992 |
+
"learning_rate": 0.00011526315789473682,
|
| 993 |
+
"loss": 0.0054,
|
| 994 |
+
"step": 1270
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 1280.0,
|
| 998 |
+
"grad_norm": 0.004704161547124386,
|
| 999 |
+
"learning_rate": 0.00011368421052631579,
|
| 1000 |
+
"loss": 0.0054,
|
| 1001 |
+
"step": 1280
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 1290.0,
|
| 1005 |
+
"grad_norm": 0.00560784712433815,
|
| 1006 |
+
"learning_rate": 0.00011210526315789472,
|
| 1007 |
+
"loss": 0.0054,
|
| 1008 |
+
"step": 1290
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 1300.0,
|
| 1012 |
+
"grad_norm": 0.003551292000338435,
|
| 1013 |
+
"learning_rate": 0.00011052631578947366,
|
| 1014 |
+
"loss": 0.0054,
|
| 1015 |
+
"step": 1300
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 1300.0,
|
| 1019 |
+
"eval_loss": 6.797518730163574,
|
| 1020 |
+
"eval_runtime": 0.0715,
|
| 1021 |
+
"eval_samples_per_second": 223.82,
|
| 1022 |
+
"eval_steps_per_second": 13.989,
|
| 1023 |
+
"step": 1300
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 1310.0,
|
| 1027 |
+
"grad_norm": 0.0029719718731939793,
|
| 1028 |
+
"learning_rate": 0.00010894736842105263,
|
| 1029 |
+
"loss": 0.0054,
|
| 1030 |
+
"step": 1310
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1320.0,
|
| 1034 |
+
"grad_norm": 0.0032612320501357317,
|
| 1035 |
+
"learning_rate": 0.00010736842105263157,
|
| 1036 |
+
"loss": 0.0054,
|
| 1037 |
+
"step": 1320
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 1330.0,
|
| 1041 |
+
"grad_norm": 0.003693740116432309,
|
| 1042 |
+
"learning_rate": 0.00010578947368421051,
|
| 1043 |
+
"loss": 0.0054,
|
| 1044 |
+
"step": 1330
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 1340.0,
|
| 1048 |
+
"grad_norm": 0.0029208508785814047,
|
| 1049 |
+
"learning_rate": 0.00010421052631578947,
|
| 1050 |
+
"loss": 0.0054,
|
| 1051 |
+
"step": 1340
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 1350.0,
|
| 1055 |
+
"grad_norm": 0.00595273170620203,
|
| 1056 |
+
"learning_rate": 0.00010263157894736841,
|
| 1057 |
+
"loss": 0.0054,
|
| 1058 |
+
"step": 1350
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1360.0,
|
| 1062 |
+
"grad_norm": 0.003722916590049863,
|
| 1063 |
+
"learning_rate": 0.00010105263157894735,
|
| 1064 |
+
"loss": 0.0054,
|
| 1065 |
+
"step": 1360
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 1370.0,
|
| 1069 |
+
"grad_norm": 0.0032132056076079607,
|
| 1070 |
+
"learning_rate": 9.94736842105263e-05,
|
| 1071 |
+
"loss": 0.0054,
|
| 1072 |
+
"step": 1370
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 1380.0,
|
| 1076 |
+
"grad_norm": 0.004051781725138426,
|
| 1077 |
+
"learning_rate": 9.789473684210526e-05,
|
| 1078 |
+
"loss": 0.0054,
|
| 1079 |
+
"step": 1380
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 1390.0,
|
| 1083 |
+
"grad_norm": 0.0027944352477788925,
|
| 1084 |
+
"learning_rate": 9.63157894736842e-05,
|
| 1085 |
+
"loss": 0.0054,
|
| 1086 |
+
"step": 1390
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 1400.0,
|
| 1090 |
+
"grad_norm": 0.003489309921860695,
|
| 1091 |
+
"learning_rate": 9.473684210526315e-05,
|
| 1092 |
+
"loss": 0.0054,
|
| 1093 |
+
"step": 1400
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 1400.0,
|
| 1097 |
+
"eval_loss": 6.807862281799316,
|
| 1098 |
+
"eval_runtime": 0.0784,
|
| 1099 |
+
"eval_samples_per_second": 204.11,
|
| 1100 |
+
"eval_steps_per_second": 12.757,
|
| 1101 |
+
"step": 1400
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1410.0,
|
| 1105 |
+
"grad_norm": 0.0035617060493677855,
|
| 1106 |
+
"learning_rate": 9.31578947368421e-05,
|
| 1107 |
+
"loss": 0.0054,
|
| 1108 |
+
"step": 1410
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1420.0,
|
| 1112 |
+
"grad_norm": 0.00412606680765748,
|
| 1113 |
+
"learning_rate": 9.157894736842104e-05,
|
| 1114 |
+
"loss": 0.0054,
|
| 1115 |
+
"step": 1420
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1430.0,
|
| 1119 |
+
"grad_norm": 0.0025130140129476786,
|
| 1120 |
+
"learning_rate": 8.999999999999999e-05,
|
| 1121 |
+
"loss": 0.0054,
|
| 1122 |
+
"step": 1430
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1440.0,
|
| 1126 |
+
"grad_norm": 0.003538177814334631,
|
| 1127 |
+
"learning_rate": 8.842105263157893e-05,
|
| 1128 |
+
"loss": 0.0054,
|
| 1129 |
+
"step": 1440
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1450.0,
|
| 1133 |
+
"grad_norm": 0.0038532898761332035,
|
| 1134 |
+
"learning_rate": 8.68421052631579e-05,
|
| 1135 |
+
"loss": 0.0054,
|
| 1136 |
+
"step": 1450
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1460.0,
|
| 1140 |
+
"grad_norm": 0.0030542940367013216,
|
| 1141 |
+
"learning_rate": 8.526315789473684e-05,
|
| 1142 |
+
"loss": 0.0054,
|
| 1143 |
+
"step": 1460
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1470.0,
|
| 1147 |
+
"grad_norm": 0.003488674759864807,
|
| 1148 |
+
"learning_rate": 8.368421052631578e-05,
|
| 1149 |
+
"loss": 0.0054,
|
| 1150 |
+
"step": 1470
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1480.0,
|
| 1154 |
+
"grad_norm": 0.0034734217915683985,
|
| 1155 |
+
"learning_rate": 8.210526315789474e-05,
|
| 1156 |
+
"loss": 0.0054,
|
| 1157 |
+
"step": 1480
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1490.0,
|
| 1161 |
+
"grad_norm": 0.00296577624976635,
|
| 1162 |
+
"learning_rate": 8.052631578947368e-05,
|
| 1163 |
+
"loss": 0.0054,
|
| 1164 |
+
"step": 1490
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1500.0,
|
| 1168 |
+
"grad_norm": 0.004232010338455439,
|
| 1169 |
+
"learning_rate": 7.894736842105262e-05,
|
| 1170 |
+
"loss": 0.0054,
|
| 1171 |
+
"step": 1500
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1500.0,
|
| 1175 |
+
"eval_loss": 6.816506862640381,
|
| 1176 |
+
"eval_runtime": 0.0566,
|
| 1177 |
+
"eval_samples_per_second": 282.528,
|
| 1178 |
+
"eval_steps_per_second": 17.658,
|
| 1179 |
+
"step": 1500
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1510.0,
|
| 1183 |
+
"grad_norm": 0.00355426874011755,
|
| 1184 |
+
"learning_rate": 7.736842105263159e-05,
|
| 1185 |
+
"loss": 0.0054,
|
| 1186 |
+
"step": 1510
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1520.0,
|
| 1190 |
+
"grad_norm": 0.0033336167689412832,
|
| 1191 |
+
"learning_rate": 7.578947368421052e-05,
|
| 1192 |
+
"loss": 0.0054,
|
| 1193 |
+
"step": 1520
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1530.0,
|
| 1197 |
+
"grad_norm": 0.0036274876911193132,
|
| 1198 |
+
"learning_rate": 7.421052631578946e-05,
|
| 1199 |
+
"loss": 0.0054,
|
| 1200 |
+
"step": 1530
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1540.0,
|
| 1204 |
+
"grad_norm": 0.004252485930919647,
|
| 1205 |
+
"learning_rate": 7.263157894736842e-05,
|
| 1206 |
+
"loss": 0.0054,
|
| 1207 |
+
"step": 1540
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1550.0,
|
| 1211 |
+
"grad_norm": 0.0027540773153305054,
|
| 1212 |
+
"learning_rate": 7.105263157894735e-05,
|
| 1213 |
+
"loss": 0.0054,
|
| 1214 |
+
"step": 1550
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1560.0,
|
| 1218 |
+
"grad_norm": 0.004094436764717102,
|
| 1219 |
+
"learning_rate": 6.947368421052631e-05,
|
| 1220 |
+
"loss": 0.0054,
|
| 1221 |
+
"step": 1560
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1570.0,
|
| 1225 |
+
"grad_norm": 0.006011336576193571,
|
| 1226 |
+
"learning_rate": 6.789473684210526e-05,
|
| 1227 |
+
"loss": 0.0054,
|
| 1228 |
+
"step": 1570
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1580.0,
|
| 1232 |
+
"grad_norm": 0.006121497601270676,
|
| 1233 |
+
"learning_rate": 6.63157894736842e-05,
|
| 1234 |
+
"loss": 0.0054,
|
| 1235 |
+
"step": 1580
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1590.0,
|
| 1239 |
+
"grad_norm": 0.00510810874402523,
|
| 1240 |
+
"learning_rate": 6.473684210526315e-05,
|
| 1241 |
+
"loss": 0.0054,
|
| 1242 |
+
"step": 1590
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1600.0,
|
| 1246 |
+
"grad_norm": 0.0036733602173626423,
|
| 1247 |
+
"learning_rate": 6.315789473684209e-05,
|
| 1248 |
+
"loss": 0.0054,
|
| 1249 |
+
"step": 1600
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1600.0,
|
| 1253 |
+
"eval_loss": 6.823525428771973,
|
| 1254 |
+
"eval_runtime": 0.0562,
|
| 1255 |
+
"eval_samples_per_second": 284.573,
|
| 1256 |
+
"eval_steps_per_second": 17.786,
|
| 1257 |
+
"step": 1600
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1610.0,
|
| 1261 |
+
"grad_norm": 0.0023464614059776068,
|
| 1262 |
+
"learning_rate": 6.157894736842104e-05,
|
| 1263 |
+
"loss": 0.0054,
|
| 1264 |
+
"step": 1610
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1620.0,
|
| 1268 |
+
"grad_norm": 0.003976115491241217,
|
| 1269 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 1270 |
+
"loss": 0.0054,
|
| 1271 |
+
"step": 1620
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1630.0,
|
| 1275 |
+
"grad_norm": 0.002651224611327052,
|
| 1276 |
+
"learning_rate": 5.842105263157894e-05,
|
| 1277 |
+
"loss": 0.0054,
|
| 1278 |
+
"step": 1630
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1640.0,
|
| 1282 |
+
"grad_norm": 0.002472205553203821,
|
| 1283 |
+
"learning_rate": 5.684210526315789e-05,
|
| 1284 |
+
"loss": 0.0054,
|
| 1285 |
+
"step": 1640
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1650.0,
|
| 1289 |
+
"grad_norm": 0.0021227358374744654,
|
| 1290 |
+
"learning_rate": 5.526315789473683e-05,
|
| 1291 |
+
"loss": 0.0054,
|
| 1292 |
+
"step": 1650
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1660.0,
|
| 1296 |
+
"grad_norm": 0.002482566749677062,
|
| 1297 |
+
"learning_rate": 5.3684210526315784e-05,
|
| 1298 |
+
"loss": 0.0054,
|
| 1299 |
+
"step": 1660
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1670.0,
|
| 1303 |
+
"grad_norm": 0.00298318755812943,
|
| 1304 |
+
"learning_rate": 5.210526315789474e-05,
|
| 1305 |
+
"loss": 0.0054,
|
| 1306 |
+
"step": 1670
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1680.0,
|
| 1310 |
+
"grad_norm": 0.002869119867682457,
|
| 1311 |
+
"learning_rate": 5.0526315789473676e-05,
|
| 1312 |
+
"loss": 0.0054,
|
| 1313 |
+
"step": 1680
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1690.0,
|
| 1317 |
+
"grad_norm": 0.0015972091350704432,
|
| 1318 |
+
"learning_rate": 4.894736842105263e-05,
|
| 1319 |
+
"loss": 0.0054,
|
| 1320 |
+
"step": 1690
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1700.0,
|
| 1324 |
+
"grad_norm": 0.0019471285631880164,
|
| 1325 |
+
"learning_rate": 4.7368421052631574e-05,
|
| 1326 |
+
"loss": 0.0054,
|
| 1327 |
+
"step": 1700
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 1700.0,
|
| 1331 |
+
"eval_loss": 6.832122802734375,
|
| 1332 |
+
"eval_runtime": 0.0565,
|
| 1333 |
+
"eval_samples_per_second": 283.107,
|
| 1334 |
+
"eval_steps_per_second": 17.694,
|
| 1335 |
+
"step": 1700
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 1710.0,
|
| 1339 |
+
"grad_norm": 0.0015862587606534362,
|
| 1340 |
+
"learning_rate": 4.578947368421052e-05,
|
| 1341 |
+
"loss": 0.0054,
|
| 1342 |
+
"step": 1710
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 1720.0,
|
| 1346 |
+
"grad_norm": 0.0018773003248497844,
|
| 1347 |
+
"learning_rate": 4.4210526315789466e-05,
|
| 1348 |
+
"loss": 0.0054,
|
| 1349 |
+
"step": 1720
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1730.0,
|
| 1353 |
+
"grad_norm": 0.0024244124069809914,
|
| 1354 |
+
"learning_rate": 4.263157894736842e-05,
|
| 1355 |
+
"loss": 0.0054,
|
| 1356 |
+
"step": 1730
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 1740.0,
|
| 1360 |
+
"grad_norm": 0.0032190545462071896,
|
| 1361 |
+
"learning_rate": 4.105263157894737e-05,
|
| 1362 |
+
"loss": 0.0054,
|
| 1363 |
+
"step": 1740
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 1750.0,
|
| 1367 |
+
"grad_norm": 0.002826133742928505,
|
| 1368 |
+
"learning_rate": 3.947368421052631e-05,
|
| 1369 |
+
"loss": 0.0054,
|
| 1370 |
+
"step": 1750
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 1760.0,
|
| 1374 |
+
"grad_norm": 0.002389115747064352,
|
| 1375 |
+
"learning_rate": 3.789473684210526e-05,
|
| 1376 |
+
"loss": 0.0054,
|
| 1377 |
+
"step": 1760
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1770.0,
|
| 1381 |
+
"grad_norm": 0.0018089566146954894,
|
| 1382 |
+
"learning_rate": 3.631578947368421e-05,
|
| 1383 |
+
"loss": 0.0054,
|
| 1384 |
+
"step": 1770
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 1780.0,
|
| 1388 |
+
"grad_norm": 0.001812565722502768,
|
| 1389 |
+
"learning_rate": 3.4736842105263153e-05,
|
| 1390 |
+
"loss": 0.0054,
|
| 1391 |
+
"step": 1780
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1790.0,
|
| 1395 |
+
"grad_norm": 0.0015028449706733227,
|
| 1396 |
+
"learning_rate": 3.31578947368421e-05,
|
| 1397 |
+
"loss": 0.0054,
|
| 1398 |
+
"step": 1790
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 1800.0,
|
| 1402 |
+
"grad_norm": 0.00215025688521564,
|
| 1403 |
+
"learning_rate": 3.1578947368421045e-05,
|
| 1404 |
+
"loss": 0.0054,
|
| 1405 |
+
"step": 1800
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 1800.0,
|
| 1409 |
+
"eval_loss": 6.83547306060791,
|
| 1410 |
+
"eval_runtime": 0.0569,
|
| 1411 |
+
"eval_samples_per_second": 281.101,
|
| 1412 |
+
"eval_steps_per_second": 17.569,
|
| 1413 |
+
"step": 1800
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1810.0,
|
| 1417 |
+
"grad_norm": 0.002039752434939146,
|
| 1418 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 1419 |
+
"loss": 0.0054,
|
| 1420 |
+
"step": 1810
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1820.0,
|
| 1424 |
+
"grad_norm": 0.0030053190421313047,
|
| 1425 |
+
"learning_rate": 2.8421052631578946e-05,
|
| 1426 |
+
"loss": 0.0054,
|
| 1427 |
+
"step": 1820
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1830.0,
|
| 1431 |
+
"grad_norm": 0.0024026986211538315,
|
| 1432 |
+
"learning_rate": 2.6842105263157892e-05,
|
| 1433 |
+
"loss": 0.0054,
|
| 1434 |
+
"step": 1830
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1840.0,
|
| 1438 |
+
"grad_norm": 0.001505296560935676,
|
| 1439 |
+
"learning_rate": 2.5263157894736838e-05,
|
| 1440 |
+
"loss": 0.0054,
|
| 1441 |
+
"step": 1840
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 1850.0,
|
| 1445 |
+
"grad_norm": 0.001797975623048842,
|
| 1446 |
+
"learning_rate": 2.3684210526315787e-05,
|
| 1447 |
+
"loss": 0.0054,
|
| 1448 |
+
"step": 1850
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1860.0,
|
| 1452 |
+
"grad_norm": 0.0021602166816592216,
|
| 1453 |
+
"learning_rate": 2.2105263157894733e-05,
|
| 1454 |
+
"loss": 0.0054,
|
| 1455 |
+
"step": 1860
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1870.0,
|
| 1459 |
+
"grad_norm": 0.0020166405010968447,
|
| 1460 |
+
"learning_rate": 2.0526315789473685e-05,
|
| 1461 |
+
"loss": 0.0054,
|
| 1462 |
+
"step": 1870
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 1880.0,
|
| 1466 |
+
"grad_norm": 0.0018515754491090775,
|
| 1467 |
+
"learning_rate": 1.894736842105263e-05,
|
| 1468 |
+
"loss": 0.0054,
|
| 1469 |
+
"step": 1880
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1890.0,
|
| 1473 |
+
"grad_norm": 0.0014232598477974534,
|
| 1474 |
+
"learning_rate": 1.7368421052631577e-05,
|
| 1475 |
+
"loss": 0.0054,
|
| 1476 |
+
"step": 1890
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 1900.0,
|
| 1480 |
+
"grad_norm": 0.0013582637766376138,
|
| 1481 |
+
"learning_rate": 1.5789473684210522e-05,
|
| 1482 |
+
"loss": 0.0054,
|
| 1483 |
+
"step": 1900
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 1900.0,
|
| 1487 |
+
"eval_loss": 6.838393688201904,
|
| 1488 |
+
"eval_runtime": 0.0566,
|
| 1489 |
+
"eval_samples_per_second": 282.616,
|
| 1490 |
+
"eval_steps_per_second": 17.663,
|
| 1491 |
+
"step": 1900
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 1910.0,
|
| 1495 |
+
"grad_norm": 0.0013303229352459311,
|
| 1496 |
+
"learning_rate": 1.4210526315789473e-05,
|
| 1497 |
+
"loss": 0.0054,
|
| 1498 |
+
"step": 1910
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 1920.0,
|
| 1502 |
+
"grad_norm": 0.0014441731618717313,
|
| 1503 |
+
"learning_rate": 1.2631578947368419e-05,
|
| 1504 |
+
"loss": 0.0054,
|
| 1505 |
+
"step": 1920
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1930.0,
|
| 1509 |
+
"grad_norm": 0.0011644844198599458,
|
| 1510 |
+
"learning_rate": 1.1052631578947366e-05,
|
| 1511 |
+
"loss": 0.0054,
|
| 1512 |
+
"step": 1930
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 1940.0,
|
| 1516 |
+
"grad_norm": 0.0011815495090559125,
|
| 1517 |
+
"learning_rate": 9.473684210526315e-06,
|
| 1518 |
+
"loss": 0.0054,
|
| 1519 |
+
"step": 1940
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 1950.0,
|
| 1523 |
+
"grad_norm": 0.0019422987243160605,
|
| 1524 |
+
"learning_rate": 7.894736842105261e-06,
|
| 1525 |
+
"loss": 0.0054,
|
| 1526 |
+
"step": 1950
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 1960.0,
|
| 1530 |
+
"grad_norm": 0.0012022488517686725,
|
| 1531 |
+
"learning_rate": 6.3157894736842095e-06,
|
| 1532 |
+
"loss": 0.0054,
|
| 1533 |
+
"step": 1960
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 1970.0,
|
| 1537 |
+
"grad_norm": 0.0013600951060652733,
|
| 1538 |
+
"learning_rate": 4.736842105263158e-06,
|
| 1539 |
+
"loss": 0.0054,
|
| 1540 |
+
"step": 1970
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1980.0,
|
| 1544 |
+
"grad_norm": 0.0014138800324872136,
|
| 1545 |
+
"learning_rate": 3.1578947368421047e-06,
|
| 1546 |
+
"loss": 0.0054,
|
| 1547 |
+
"step": 1980
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 1990.0,
|
| 1551 |
+
"grad_norm": 0.0012338694650679827,
|
| 1552 |
+
"learning_rate": 1.5789473684210524e-06,
|
| 1553 |
+
"loss": 0.0054,
|
| 1554 |
+
"step": 1990
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 2000.0,
|
| 1558 |
+
"grad_norm": 0.0009407839388586581,
|
| 1559 |
+
"learning_rate": 0.0,
|
| 1560 |
+
"loss": 0.0054,
|
| 1561 |
+
"step": 2000
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 2000.0,
|
| 1565 |
+
"eval_loss": 6.840120792388916,
|
| 1566 |
+
"eval_runtime": 0.0569,
|
| 1567 |
+
"eval_samples_per_second": 280.972,
|
| 1568 |
+
"eval_steps_per_second": 17.561,
|
| 1569 |
+
"step": 2000
|
| 1570 |
+
}
|
| 1571 |
+
],
|
| 1572 |
+
"logging_steps": 10,
|
| 1573 |
+
"max_steps": 2000,
|
| 1574 |
+
"num_input_tokens_seen": 0,
|
| 1575 |
+
"num_train_epochs": 2000,
|
| 1576 |
+
"save_steps": 100,
|
| 1577 |
+
"stateful_callbacks": {
|
| 1578 |
+
"TrainerControl": {
|
| 1579 |
+
"args": {
|
| 1580 |
+
"should_epoch_stop": false,
|
| 1581 |
+
"should_evaluate": false,
|
| 1582 |
+
"should_log": false,
|
| 1583 |
+
"should_save": true,
|
| 1584 |
+
"should_training_stop": true
|
| 1585 |
+
},
|
| 1586 |
+
"attributes": {}
|
| 1587 |
+
}
|
| 1588 |
+
},
|
| 1589 |
+
"total_flos": 1.045476409344e+16,
|
| 1590 |
+
"train_batch_size": 30,
|
| 1591 |
+
"trial_name": null,
|
| 1592 |
+
"trial_params": null
|
| 1593 |
+
}
|
instruct-finetuning/base/LLaMmlein_120M/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:49b230fcb8bb33325a80b3b5f9018ab82781c698b9f593e510fec637714f64d6
|
| 3 |
+
size 5496
|
instruct-finetuning/base/german-gpt2/README.md
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: dbmdz/german-gpt2
|
| 3 |
+
library_name: peft
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# Model Card for Model ID
|
| 7 |
+
|
| 8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
### Model Description
|
| 15 |
+
|
| 16 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [More Information Needed]
|
| 21 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 22 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 23 |
+
- **Model type:** [More Information Needed]
|
| 24 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 25 |
+
- **License:** [More Information Needed]
|
| 26 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 27 |
+
|
| 28 |
+
### Model Sources [optional]
|
| 29 |
+
|
| 30 |
+
<!-- Provide the basic links for the model. -->
|
| 31 |
+
|
| 32 |
+
- **Repository:** [More Information Needed]
|
| 33 |
+
- **Paper [optional]:** [More Information Needed]
|
| 34 |
+
- **Demo [optional]:** [More Information Needed]
|
| 35 |
+
|
| 36 |
+
## Uses
|
| 37 |
+
|
| 38 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 39 |
+
|
| 40 |
+
### Direct Use
|
| 41 |
+
|
| 42 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 43 |
+
|
| 44 |
+
[More Information Needed]
|
| 45 |
+
|
| 46 |
+
### Downstream Use [optional]
|
| 47 |
+
|
| 48 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 49 |
+
|
| 50 |
+
[More Information Needed]
|
| 51 |
+
|
| 52 |
+
### Out-of-Scope Use
|
| 53 |
+
|
| 54 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 55 |
+
|
| 56 |
+
[More Information Needed]
|
| 57 |
+
|
| 58 |
+
## Bias, Risks, and Limitations
|
| 59 |
+
|
| 60 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 61 |
+
|
| 62 |
+
[More Information Needed]
|
| 63 |
+
|
| 64 |
+
### Recommendations
|
| 65 |
+
|
| 66 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 67 |
+
|
| 68 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 69 |
+
|
| 70 |
+
## How to Get Started with the Model
|
| 71 |
+
|
| 72 |
+
Use the code below to get started with the model.
|
| 73 |
+
|
| 74 |
+
[More Information Needed]
|
| 75 |
+
|
| 76 |
+
## Training Details
|
| 77 |
+
|
| 78 |
+
### Training Data
|
| 79 |
+
|
| 80 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 81 |
+
|
| 82 |
+
[More Information Needed]
|
| 83 |
+
|
| 84 |
+
### Training Procedure
|
| 85 |
+
|
| 86 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 87 |
+
|
| 88 |
+
#### Preprocessing [optional]
|
| 89 |
+
|
| 90 |
+
[More Information Needed]
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
#### Training Hyperparameters
|
| 94 |
+
|
| 95 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 96 |
+
|
| 97 |
+
#### Speeds, Sizes, Times [optional]
|
| 98 |
+
|
| 99 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 100 |
+
|
| 101 |
+
[More Information Needed]
|
| 102 |
+
|
| 103 |
+
## Evaluation
|
| 104 |
+
|
| 105 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 106 |
+
|
| 107 |
+
### Testing Data, Factors & Metrics
|
| 108 |
+
|
| 109 |
+
#### Testing Data
|
| 110 |
+
|
| 111 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 112 |
+
|
| 113 |
+
[More Information Needed]
|
| 114 |
+
|
| 115 |
+
#### Factors
|
| 116 |
+
|
| 117 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 118 |
+
|
| 119 |
+
[More Information Needed]
|
| 120 |
+
|
| 121 |
+
#### Metrics
|
| 122 |
+
|
| 123 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 124 |
+
|
| 125 |
+
[More Information Needed]
|
| 126 |
+
|
| 127 |
+
### Results
|
| 128 |
+
|
| 129 |
+
[More Information Needed]
|
| 130 |
+
|
| 131 |
+
#### Summary
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
|
| 135 |
+
## Model Examination [optional]
|
| 136 |
+
|
| 137 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 138 |
+
|
| 139 |
+
[More Information Needed]
|
| 140 |
+
|
| 141 |
+
## Environmental Impact
|
| 142 |
+
|
| 143 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 144 |
+
|
| 145 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 146 |
+
|
| 147 |
+
- **Hardware Type:** [More Information Needed]
|
| 148 |
+
- **Hours used:** [More Information Needed]
|
| 149 |
+
- **Cloud Provider:** [More Information Needed]
|
| 150 |
+
- **Compute Region:** [More Information Needed]
|
| 151 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 152 |
+
|
| 153 |
+
## Technical Specifications [optional]
|
| 154 |
+
|
| 155 |
+
### Model Architecture and Objective
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
### Compute Infrastructure
|
| 160 |
+
|
| 161 |
+
[More Information Needed]
|
| 162 |
+
|
| 163 |
+
#### Hardware
|
| 164 |
+
|
| 165 |
+
[More Information Needed]
|
| 166 |
+
|
| 167 |
+
#### Software
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
## Citation [optional]
|
| 172 |
+
|
| 173 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 174 |
+
|
| 175 |
+
**BibTeX:**
|
| 176 |
+
|
| 177 |
+
[More Information Needed]
|
| 178 |
+
|
| 179 |
+
**APA:**
|
| 180 |
+
|
| 181 |
+
[More Information Needed]
|
| 182 |
+
|
| 183 |
+
## Glossary [optional]
|
| 184 |
+
|
| 185 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 186 |
+
|
| 187 |
+
[More Information Needed]
|
| 188 |
+
|
| 189 |
+
## More Information [optional]
|
| 190 |
+
|
| 191 |
+
[More Information Needed]
|
| 192 |
+
|
| 193 |
+
## Model Card Authors [optional]
|
| 194 |
+
|
| 195 |
+
[More Information Needed]
|
| 196 |
+
|
| 197 |
+
## Model Card Contact
|
| 198 |
+
|
| 199 |
+
[More Information Needed]
|
| 200 |
+
### Framework versions
|
| 201 |
+
|
| 202 |
+
- PEFT 0.13.2
|
instruct-finetuning/base/german-gpt2/adapter_config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "dbmdz/german-gpt2",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layer_replication": null,
|
| 10 |
+
"layers_pattern": null,
|
| 11 |
+
"layers_to_transform": null,
|
| 12 |
+
"loftq_config": {},
|
| 13 |
+
"lora_alpha": 4,
|
| 14 |
+
"lora_dropout": 0.1,
|
| 15 |
+
"megatron_config": null,
|
| 16 |
+
"megatron_core": "megatron.core",
|
| 17 |
+
"modules_to_save": null,
|
| 18 |
+
"peft_type": "LORA",
|
| 19 |
+
"r": 32,
|
| 20 |
+
"rank_pattern": {},
|
| 21 |
+
"revision": null,
|
| 22 |
+
"target_modules": [
|
| 23 |
+
"k_proj",
|
| 24 |
+
"down_proj",
|
| 25 |
+
"v_proj",
|
| 26 |
+
"lm_head",
|
| 27 |
+
"gate_proj",
|
| 28 |
+
"q_proj",
|
| 29 |
+
"up_proj",
|
| 30 |
+
"o_proj"
|
| 31 |
+
],
|
| 32 |
+
"task_type": "CAUSAL_LM",
|
| 33 |
+
"use_dora": false,
|
| 34 |
+
"use_rslora": false
|
| 35 |
+
}
|
instruct-finetuning/base/german-gpt2/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c02604806e39e6f29920141a48c20f0df2a13c16f1755d10617a733316a23f3
|
| 3 |
+
size 160949880
|
instruct-finetuning/base/german-gpt2/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<|endoftext|>": 50265
|
| 3 |
+
}
|
instruct-finetuning/base/german-gpt2/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
instruct-finetuning/base/german-gpt2/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84356959f433ca56d0418c13cc0eb6e657c4ebac43251dced6e86ba9a281708d
|
| 3 |
+
size 13067612
|
instruct-finetuning/base/german-gpt2/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e862fec9943ccd0a1ec4c4847bbc61f171f1c24a21f76d7a53e951110d236b4
|
| 3 |
+
size 14244
|