Upload folder using huggingface_hub
Browse files
slots/8/checkpoint-5992/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"TwinyForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"hidden_dropout": 0.0,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 3072,
|
| 11 |
+
"max_position_embeddings": 128,
|
| 12 |
+
"model_type": "twiny",
|
| 13 |
+
"neftune_alpha": 0.0,
|
| 14 |
+
"num_attention_heads": 12,
|
| 15 |
+
"num_hidden_layers": 3,
|
| 16 |
+
"num_key_value_heads": 3,
|
| 17 |
+
"qk_norm": true,
|
| 18 |
+
"rezero_init": 1.0,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 10000.0,
|
| 21 |
+
"transformers_version": "5.0.0",
|
| 22 |
+
"use_cache": false,
|
| 23 |
+
"vocab_size": 32000
|
| 24 |
+
}
|
slots/8/checkpoint-5992/trainer_state.json
ADDED
|
@@ -0,0 +1,2134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.1214734978611898,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 5992,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0001336931908386741,
|
| 14 |
+
"grad_norm": Infinity,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 129.80032348632812,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.002673863816773482,
|
| 21 |
+
"grad_norm": 63.64365768432617,
|
| 22 |
+
"learning_rate": 4.999995392022967e-05,
|
| 23 |
+
"loss": 63.88374408922697,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.005347727633546964,
|
| 28 |
+
"grad_norm": 24.627853393554688,
|
| 29 |
+
"learning_rate": 4.999976672145381e-05,
|
| 30 |
+
"loss": 12.65963363647461,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.008021591450320446,
|
| 35 |
+
"grad_norm": 14.29983901977539,
|
| 36 |
+
"learning_rate": 4.999943552476422e-05,
|
| 37 |
+
"loss": 5.90204963684082,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.010695455267093928,
|
| 42 |
+
"grad_norm": 15.690323829650879,
|
| 43 |
+
"learning_rate": 4.999896033206858e-05,
|
| 44 |
+
"loss": 3.9918922424316405,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.01336931908386741,
|
| 49 |
+
"grad_norm": 31.583160400390625,
|
| 50 |
+
"learning_rate": 4.999834114610398e-05,
|
| 51 |
+
"loss": 2.9675426483154297,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.01604318290064089,
|
| 56 |
+
"grad_norm": 13.034649848937988,
|
| 57 |
+
"learning_rate": 4.999757797043691e-05,
|
| 58 |
+
"loss": 2.725296401977539,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.018717046717414372,
|
| 63 |
+
"grad_norm": 8.362203598022461,
|
| 64 |
+
"learning_rate": 4.999667080946324e-05,
|
| 65 |
+
"loss": 2.2478992462158205,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.021390910534187856,
|
| 70 |
+
"grad_norm": 8.726786613464355,
|
| 71 |
+
"learning_rate": 4.999561966840821e-05,
|
| 72 |
+
"loss": 1.8447845458984375,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.024064774350961337,
|
| 77 |
+
"grad_norm": 10.092752456665039,
|
| 78 |
+
"learning_rate": 4.9994424553326335e-05,
|
| 79 |
+
"loss": 1.5611843109130858,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02673863816773482,
|
| 84 |
+
"grad_norm": 9.090085983276367,
|
| 85 |
+
"learning_rate": 4.999308547110146e-05,
|
| 86 |
+
"loss": 1.520334815979004,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.029412501984508302,
|
| 91 |
+
"grad_norm": 9.668124198913574,
|
| 92 |
+
"learning_rate": 4.999160242944665e-05,
|
| 93 |
+
"loss": 1.2818055152893066,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.03208636580128178,
|
| 98 |
+
"grad_norm": 9.182533264160156,
|
| 99 |
+
"learning_rate": 4.998997543690418e-05,
|
| 100 |
+
"loss": 1.0428407669067383,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.03476022961805526,
|
| 105 |
+
"grad_norm": 5.745838165283203,
|
| 106 |
+
"learning_rate": 4.998820450284549e-05,
|
| 107 |
+
"loss": 1.2343652725219727,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.037434093434828744,
|
| 112 |
+
"grad_norm": 8.651643753051758,
|
| 113 |
+
"learning_rate": 4.99862896374711e-05,
|
| 114 |
+
"loss": 0.8859601020812988,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04010795725160223,
|
| 119 |
+
"grad_norm": 10.765266418457031,
|
| 120 |
+
"learning_rate": 4.998423085181056e-05,
|
| 121 |
+
"loss": 0.989600658416748,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.04278182106837571,
|
| 126 |
+
"grad_norm": 6.092499256134033,
|
| 127 |
+
"learning_rate": 4.998202815772245e-05,
|
| 128 |
+
"loss": 0.7189463615417481,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.04545568488514919,
|
| 133 |
+
"grad_norm": 6.352876663208008,
|
| 134 |
+
"learning_rate": 4.9979681567894195e-05,
|
| 135 |
+
"loss": 0.7489545345306396,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.048129548701922674,
|
| 140 |
+
"grad_norm": 4.620656490325928,
|
| 141 |
+
"learning_rate": 4.997719109584209e-05,
|
| 142 |
+
"loss": 0.7381401538848877,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.050803412518696155,
|
| 147 |
+
"grad_norm": 7.796917915344238,
|
| 148 |
+
"learning_rate": 4.997455675591119e-05,
|
| 149 |
+
"loss": 0.5687405109405518,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.05347727633546964,
|
| 154 |
+
"grad_norm": 2.837172508239746,
|
| 155 |
+
"learning_rate": 4.9971778563275204e-05,
|
| 156 |
+
"loss": 0.5686865329742432,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.05615114015224312,
|
| 161 |
+
"grad_norm": 3.3103690147399902,
|
| 162 |
+
"learning_rate": 4.9968856533936436e-05,
|
| 163 |
+
"loss": 0.625730562210083,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.058825003969016604,
|
| 168 |
+
"grad_norm": 3.5682132244110107,
|
| 169 |
+
"learning_rate": 4.99657906847257e-05,
|
| 170 |
+
"loss": 0.6125466346740722,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.061498867785790085,
|
| 175 |
+
"grad_norm": 5.63640832901001,
|
| 176 |
+
"learning_rate": 4.996258103330218e-05,
|
| 177 |
+
"loss": 0.6182214260101319,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.06417273160256357,
|
| 182 |
+
"grad_norm": 4.698945999145508,
|
| 183 |
+
"learning_rate": 4.995922759815339e-05,
|
| 184 |
+
"loss": 0.43828091621398924,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.06684659541933705,
|
| 189 |
+
"grad_norm": 2.1976189613342285,
|
| 190 |
+
"learning_rate": 4.995573039859501e-05,
|
| 191 |
+
"loss": 0.4459230899810791,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.06952045923611053,
|
| 196 |
+
"grad_norm": 3.8809523582458496,
|
| 197 |
+
"learning_rate": 4.995208945477081e-05,
|
| 198 |
+
"loss": 0.3821882963180542,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.07219432305288401,
|
| 203 |
+
"grad_norm": 3.75144100189209,
|
| 204 |
+
"learning_rate": 4.994830478765251e-05,
|
| 205 |
+
"loss": 0.5800807476043701,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.07486818686965749,
|
| 210 |
+
"grad_norm": 3.0038585662841797,
|
| 211 |
+
"learning_rate": 4.9944376419039684e-05,
|
| 212 |
+
"loss": 0.3928264617919922,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07754205068643098,
|
| 217 |
+
"grad_norm": 3.614591598510742,
|
| 218 |
+
"learning_rate": 4.994030437155961e-05,
|
| 219 |
+
"loss": 0.48637890815734863,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.08021591450320446,
|
| 224 |
+
"grad_norm": 4.143443584442139,
|
| 225 |
+
"learning_rate": 4.993608866866718e-05,
|
| 226 |
+
"loss": 0.3650153160095215,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.08288977831997794,
|
| 231 |
+
"grad_norm": 6.692712783813477,
|
| 232 |
+
"learning_rate": 4.993172933464471e-05,
|
| 233 |
+
"loss": 0.3677916288375854,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.08556364213675142,
|
| 238 |
+
"grad_norm": 8.383441925048828,
|
| 239 |
+
"learning_rate": 4.9927226394601815e-05,
|
| 240 |
+
"loss": 0.3399480104446411,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.0882375059535249,
|
| 245 |
+
"grad_norm": 5.566338062286377,
|
| 246 |
+
"learning_rate": 4.992257987447532e-05,
|
| 247 |
+
"loss": 0.28104052543640134,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.09091136977029839,
|
| 252 |
+
"grad_norm": 3.1196420192718506,
|
| 253 |
+
"learning_rate": 4.991778980102904e-05,
|
| 254 |
+
"loss": 0.351950478553772,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.09358523358707187,
|
| 259 |
+
"grad_norm": 3.47979736328125,
|
| 260 |
+
"learning_rate": 4.9912856201853644e-05,
|
| 261 |
+
"loss": 0.27501535415649414,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.09625909740384535,
|
| 266 |
+
"grad_norm": 5.446717262268066,
|
| 267 |
+
"learning_rate": 4.990777910536653e-05,
|
| 268 |
+
"loss": 0.2651593923568726,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.09893296122061883,
|
| 273 |
+
"grad_norm": 7.6145339012146,
|
| 274 |
+
"learning_rate": 4.990255854081161e-05,
|
| 275 |
+
"loss": 0.35140380859375,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.10160682503739231,
|
| 280 |
+
"grad_norm": 8.445616722106934,
|
| 281 |
+
"learning_rate": 4.989719453825918e-05,
|
| 282 |
+
"loss": 0.2961219072341919,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.10428068885416579,
|
| 287 |
+
"grad_norm": 6.339537620544434,
|
| 288 |
+
"learning_rate": 4.9891687128605744e-05,
|
| 289 |
+
"loss": 0.24962289333343507,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.10695455267093928,
|
| 294 |
+
"grad_norm": 3.3369436264038086,
|
| 295 |
+
"learning_rate": 4.988603634357383e-05,
|
| 296 |
+
"loss": 0.2124847412109375,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.10962841648771277,
|
| 301 |
+
"grad_norm": 2.2909045219421387,
|
| 302 |
+
"learning_rate": 4.988024221571177e-05,
|
| 303 |
+
"loss": 0.24679112434387207,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.11230228030448625,
|
| 308 |
+
"grad_norm": 3.1149911880493164,
|
| 309 |
+
"learning_rate": 4.9874304778393574e-05,
|
| 310 |
+
"loss": 0.22161397933959961,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.11497614412125973,
|
| 315 |
+
"grad_norm": 14.802160263061523,
|
| 316 |
+
"learning_rate": 4.9868224065818706e-05,
|
| 317 |
+
"loss": 0.2623537302017212,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.11765000793803321,
|
| 322 |
+
"grad_norm": 5.586325168609619,
|
| 323 |
+
"learning_rate": 4.98620001130119e-05,
|
| 324 |
+
"loss": 0.3560942649841309,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.12032387175480669,
|
| 329 |
+
"grad_norm": 3.390017032623291,
|
| 330 |
+
"learning_rate": 4.9855632955822916e-05,
|
| 331 |
+
"loss": 0.16934787034988402,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.12299773557158017,
|
| 336 |
+
"grad_norm": 6.070940971374512,
|
| 337 |
+
"learning_rate": 4.984912263092641e-05,
|
| 338 |
+
"loss": 0.2131197214126587,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.12567159938835365,
|
| 343 |
+
"grad_norm": 1.4912281036376953,
|
| 344 |
+
"learning_rate": 4.984246917582166e-05,
|
| 345 |
+
"loss": 0.25128653049468996,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.12834546320512713,
|
| 350 |
+
"grad_norm": 7.000472545623779,
|
| 351 |
+
"learning_rate": 4.9835672628832366e-05,
|
| 352 |
+
"loss": 0.2653592586517334,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.1310193270219006,
|
| 357 |
+
"grad_norm": 5.427223205566406,
|
| 358 |
+
"learning_rate": 4.9828733029106434e-05,
|
| 359 |
+
"loss": 0.1653295636177063,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.1336931908386741,
|
| 364 |
+
"grad_norm": 1.9502102136611938,
|
| 365 |
+
"learning_rate": 4.982165041661575e-05,
|
| 366 |
+
"loss": 0.2250870943069458,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.13636705465544757,
|
| 371 |
+
"grad_norm": 0.6216259598731995,
|
| 372 |
+
"learning_rate": 4.981442483215595e-05,
|
| 373 |
+
"loss": 0.18943849802017212,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.13904091847222105,
|
| 378 |
+
"grad_norm": 2.3363687992095947,
|
| 379 |
+
"learning_rate": 4.98070563173462e-05,
|
| 380 |
+
"loss": 0.1673592209815979,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.14171478228899453,
|
| 385 |
+
"grad_norm": 1.040717601776123,
|
| 386 |
+
"learning_rate": 4.979954491462892e-05,
|
| 387 |
+
"loss": 0.2113173007965088,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.14438864610576801,
|
| 392 |
+
"grad_norm": 2.735522747039795,
|
| 393 |
+
"learning_rate": 4.979189066726955e-05,
|
| 394 |
+
"loss": 0.17504971027374266,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.1470625099225415,
|
| 399 |
+
"grad_norm": 4.701151371002197,
|
| 400 |
+
"learning_rate": 4.978409361935636e-05,
|
| 401 |
+
"loss": 0.15881222486495972,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.14973637373931498,
|
| 406 |
+
"grad_norm": 2.735919237136841,
|
| 407 |
+
"learning_rate": 4.9776153815800075e-05,
|
| 408 |
+
"loss": 0.14044179916381835,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.15241023755608848,
|
| 413 |
+
"grad_norm": 3.5479538440704346,
|
| 414 |
+
"learning_rate": 4.976807130233375e-05,
|
| 415 |
+
"loss": 0.18565714359283447,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.15508410137286197,
|
| 420 |
+
"grad_norm": 3.2167458534240723,
|
| 421 |
+
"learning_rate": 4.975984612551243e-05,
|
| 422 |
+
"loss": 0.13236271142959594,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.15775796518963545,
|
| 427 |
+
"grad_norm": 1.0206760168075562,
|
| 428 |
+
"learning_rate": 4.975147833271288e-05,
|
| 429 |
+
"loss": 0.19124728441238403,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.16043182900640893,
|
| 434 |
+
"grad_norm": 4.194457530975342,
|
| 435 |
+
"learning_rate": 4.9742967972133335e-05,
|
| 436 |
+
"loss": 0.144741427898407,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.1631056928231824,
|
| 441 |
+
"grad_norm": 3.0225746631622314,
|
| 442 |
+
"learning_rate": 4.973431509279323e-05,
|
| 443 |
+
"loss": 0.1374324679374695,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.1657795566399559,
|
| 448 |
+
"grad_norm": 4.243523120880127,
|
| 449 |
+
"learning_rate": 4.972551974453287e-05,
|
| 450 |
+
"loss": 0.13663809299468993,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.16845342045672937,
|
| 455 |
+
"grad_norm": 2.4990086555480957,
|
| 456 |
+
"learning_rate": 4.971658197801322e-05,
|
| 457 |
+
"loss": 0.16817957162857056,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.17112728427350285,
|
| 462 |
+
"grad_norm": 4.983982563018799,
|
| 463 |
+
"learning_rate": 4.9707501844715554e-05,
|
| 464 |
+
"loss": 0.13795313835144044,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.17380114809027633,
|
| 469 |
+
"grad_norm": 3.6780316829681396,
|
| 470 |
+
"learning_rate": 4.969827939694115e-05,
|
| 471 |
+
"loss": 0.1637880802154541,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.1764750119070498,
|
| 476 |
+
"grad_norm": 0.7950732707977295,
|
| 477 |
+
"learning_rate": 4.968891468781105e-05,
|
| 478 |
+
"loss": 0.10979138612747193,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.1791488757238233,
|
| 483 |
+
"grad_norm": 1.2414121627807617,
|
| 484 |
+
"learning_rate": 4.967940777126569e-05,
|
| 485 |
+
"loss": 0.13692171573638917,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.18182273954059677,
|
| 490 |
+
"grad_norm": 2.1383633613586426,
|
| 491 |
+
"learning_rate": 4.9669758702064636e-05,
|
| 492 |
+
"loss": 0.07821698188781738,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.18449660335737025,
|
| 497 |
+
"grad_norm": 5.061275959014893,
|
| 498 |
+
"learning_rate": 4.965996753578623e-05,
|
| 499 |
+
"loss": 0.19053516387939454,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.18717046717414373,
|
| 504 |
+
"grad_norm": 6.151792049407959,
|
| 505 |
+
"learning_rate": 4.9650034328827305e-05,
|
| 506 |
+
"loss": 0.11360721588134766,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.18984433099091721,
|
| 511 |
+
"grad_norm": 1.0604305267333984,
|
| 512 |
+
"learning_rate": 4.963995913840284e-05,
|
| 513 |
+
"loss": 0.13138024806976317,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.1925181948076907,
|
| 518 |
+
"grad_norm": 1.7159489393234253,
|
| 519 |
+
"learning_rate": 4.9629742022545623e-05,
|
| 520 |
+
"loss": 0.08657677173614502,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.19519205862446418,
|
| 525 |
+
"grad_norm": 2.4207754135131836,
|
| 526 |
+
"learning_rate": 4.961938304010595e-05,
|
| 527 |
+
"loss": 0.10309149026870727,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.19786592244123766,
|
| 532 |
+
"grad_norm": 1.532060146331787,
|
| 533 |
+
"learning_rate": 4.9608882250751245e-05,
|
| 534 |
+
"loss": 0.13628544807434081,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.20053978625801114,
|
| 539 |
+
"grad_norm": 6.409943580627441,
|
| 540 |
+
"learning_rate": 4.959823971496574e-05,
|
| 541 |
+
"loss": 0.10584845542907714,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.20321365007478462,
|
| 546 |
+
"grad_norm": 2.452012538909912,
|
| 547 |
+
"learning_rate": 4.9587455494050136e-05,
|
| 548 |
+
"loss": 0.06506187915802002,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.2058875138915581,
|
| 553 |
+
"grad_norm": 5.3016533851623535,
|
| 554 |
+
"learning_rate": 4.9576529650121214e-05,
|
| 555 |
+
"loss": 0.11848526000976563,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.20856137770833158,
|
| 560 |
+
"grad_norm": 4.341775894165039,
|
| 561 |
+
"learning_rate": 4.956546224611152e-05,
|
| 562 |
+
"loss": 0.11318533420562744,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.21123524152510506,
|
| 567 |
+
"grad_norm": 1.9056169986724854,
|
| 568 |
+
"learning_rate": 4.9554253345768965e-05,
|
| 569 |
+
"loss": 0.12768398523330687,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.21390910534187857,
|
| 574 |
+
"grad_norm": 1.8939746618270874,
|
| 575 |
+
"learning_rate": 4.9542903013656486e-05,
|
| 576 |
+
"loss": 0.10782338380813598,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.21658296915865205,
|
| 581 |
+
"grad_norm": 8.53671932220459,
|
| 582 |
+
"learning_rate": 4.9531411315151654e-05,
|
| 583 |
+
"loss": 0.1733921766281128,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.21925683297542553,
|
| 588 |
+
"grad_norm": 2.0152978897094727,
|
| 589 |
+
"learning_rate": 4.951977831644632e-05,
|
| 590 |
+
"loss": 0.11197054386138916,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.221930696792199,
|
| 595 |
+
"grad_norm": 3.8422367572784424,
|
| 596 |
+
"learning_rate": 4.95080040845462e-05,
|
| 597 |
+
"loss": 0.11441781520843505,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.2246045606089725,
|
| 602 |
+
"grad_norm": 1.819858193397522,
|
| 603 |
+
"learning_rate": 4.949608868727053e-05,
|
| 604 |
+
"loss": 0.11403474807739258,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.22727842442574597,
|
| 609 |
+
"grad_norm": 7.45100212097168,
|
| 610 |
+
"learning_rate": 4.948403219325163e-05,
|
| 611 |
+
"loss": 0.13117753267288207,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.22995228824251945,
|
| 616 |
+
"grad_norm": 0.6526040434837341,
|
| 617 |
+
"learning_rate": 4.947183467193456e-05,
|
| 618 |
+
"loss": 0.07524924874305725,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.23262615205929293,
|
| 623 |
+
"grad_norm": 3.814746856689453,
|
| 624 |
+
"learning_rate": 4.945949619357668e-05,
|
| 625 |
+
"loss": 0.07659345269203185,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.23530001587606642,
|
| 630 |
+
"grad_norm": 2.373124122619629,
|
| 631 |
+
"learning_rate": 4.944701682924726e-05,
|
| 632 |
+
"loss": 0.1147496223449707,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.2379738796928399,
|
| 637 |
+
"grad_norm": 0.11161285638809204,
|
| 638 |
+
"learning_rate": 4.943439665082707e-05,
|
| 639 |
+
"loss": 0.07256829738616943,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.24064774350961338,
|
| 644 |
+
"grad_norm": 0.45990192890167236,
|
| 645 |
+
"learning_rate": 4.942163573100794e-05,
|
| 646 |
+
"loss": 0.07726740837097168,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.24332160732638686,
|
| 651 |
+
"grad_norm": 4.2301926612854,
|
| 652 |
+
"learning_rate": 4.940873414329242e-05,
|
| 653 |
+
"loss": 0.09349535703659058,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.24599547114316034,
|
| 658 |
+
"grad_norm": 2.442178726196289,
|
| 659 |
+
"learning_rate": 4.939569196199325e-05,
|
| 660 |
+
"loss": 0.12413722276687622,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.24866933495993382,
|
| 665 |
+
"grad_norm": 2.523683786392212,
|
| 666 |
+
"learning_rate": 4.938250926223302e-05,
|
| 667 |
+
"loss": 0.08566288352012634,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.2513431987767073,
|
| 672 |
+
"grad_norm": 3.511075258255005,
|
| 673 |
+
"learning_rate": 4.936918611994368e-05,
|
| 674 |
+
"loss": 0.08007702231407166,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.2540170625934808,
|
| 679 |
+
"grad_norm": 6.254627704620361,
|
| 680 |
+
"learning_rate": 4.935572261186614e-05,
|
| 681 |
+
"loss": 0.10983954668045044,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.25669092641025426,
|
| 686 |
+
"grad_norm": 1.5211899280548096,
|
| 687 |
+
"learning_rate": 4.934211881554981e-05,
|
| 688 |
+
"loss": 0.09120344519615173,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.25936479022702774,
|
| 693 |
+
"grad_norm": 2.5893588066101074,
|
| 694 |
+
"learning_rate": 4.932837480935214e-05,
|
| 695 |
+
"loss": 0.08754412531852722,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.2620386540438012,
|
| 700 |
+
"grad_norm": 6.878556251525879,
|
| 701 |
+
"learning_rate": 4.931449067243821e-05,
|
| 702 |
+
"loss": 0.08636274933815002,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.2647125178605747,
|
| 707 |
+
"grad_norm": 2.9078798294067383,
|
| 708 |
+
"learning_rate": 4.9300466484780226e-05,
|
| 709 |
+
"loss": 0.09582929015159607,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.2673863816773482,
|
| 714 |
+
"grad_norm": 3.391852855682373,
|
| 715 |
+
"learning_rate": 4.92863023271571e-05,
|
| 716 |
+
"loss": 0.0850919783115387,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.27006024549412166,
|
| 721 |
+
"grad_norm": 5.522103309631348,
|
| 722 |
+
"learning_rate": 4.927199828115395e-05,
|
| 723 |
+
"loss": 0.050999772548675534,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.27273410931089515,
|
| 728 |
+
"grad_norm": 0.90350741147995,
|
| 729 |
+
"learning_rate": 4.925755442916167e-05,
|
| 730 |
+
"loss": 0.10100446939468384,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.2754079731276686,
|
| 735 |
+
"grad_norm": 1.602030634880066,
|
| 736 |
+
"learning_rate": 4.924297085437641e-05,
|
| 737 |
+
"loss": 0.0468633770942688,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.2780818369444421,
|
| 742 |
+
"grad_norm": 1.5823460817337036,
|
| 743 |
+
"learning_rate": 4.922824764079913e-05,
|
| 744 |
+
"loss": 0.06786358952522278,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.2807557007612156,
|
| 749 |
+
"grad_norm": 1.6624343395233154,
|
| 750 |
+
"learning_rate": 4.92133848732351e-05,
|
| 751 |
+
"loss": 0.05772828459739685,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.28342956457798907,
|
| 756 |
+
"grad_norm": 0.947078287601471,
|
| 757 |
+
"learning_rate": 4.9198382637293424e-05,
|
| 758 |
+
"loss": 0.08012173175811768,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.28610342839476255,
|
| 763 |
+
"grad_norm": 0.2919924259185791,
|
| 764 |
+
"learning_rate": 4.918324101938653e-05,
|
| 765 |
+
"loss": 0.1208539366722107,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.28877729221153603,
|
| 770 |
+
"grad_norm": 9.258247375488281,
|
| 771 |
+
"learning_rate": 4.916796010672969e-05,
|
| 772 |
+
"loss": 0.10037034749984741,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.2914511560283095,
|
| 777 |
+
"grad_norm": 4.0920491218566895,
|
| 778 |
+
"learning_rate": 4.915253998734051e-05,
|
| 779 |
+
"loss": 0.061488878726959226,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.294125019845083,
|
| 784 |
+
"grad_norm": 6.1126627922058105,
|
| 785 |
+
"learning_rate": 4.913698075003841e-05,
|
| 786 |
+
"loss": 0.0862967312335968,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.29679888366185647,
|
| 791 |
+
"grad_norm": 2.585484743118286,
|
| 792 |
+
"learning_rate": 4.912128248444414e-05,
|
| 793 |
+
"loss": 0.05393874645233154,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.29947274747862995,
|
| 798 |
+
"grad_norm": 6.944481372833252,
|
| 799 |
+
"learning_rate": 4.9105445280979256e-05,
|
| 800 |
+
"loss": 0.08570566773414612,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.30214661129540343,
|
| 805 |
+
"grad_norm": 1.3824089765548706,
|
| 806 |
+
"learning_rate": 4.908946923086556e-05,
|
| 807 |
+
"loss": 0.09689127206802368,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.30482047511217697,
|
| 812 |
+
"grad_norm": 3.4861342906951904,
|
| 813 |
+
"learning_rate": 4.907335442612464e-05,
|
| 814 |
+
"loss": 0.12550976276397705,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.30749433892895045,
|
| 819 |
+
"grad_norm": 3.668980121612549,
|
| 820 |
+
"learning_rate": 4.905710095957728e-05,
|
| 821 |
+
"loss": 0.09089353680610657,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.31016820274572393,
|
| 826 |
+
"grad_norm": 1.093095064163208,
|
| 827 |
+
"learning_rate": 4.904070892484298e-05,
|
| 828 |
+
"loss": 0.03925192356109619,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.3128420665624974,
|
| 833 |
+
"grad_norm": 0.8169485926628113,
|
| 834 |
+
"learning_rate": 4.9024178416339364e-05,
|
| 835 |
+
"loss": 0.0979581356048584,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.3155159303792709,
|
| 840 |
+
"grad_norm": 1.892451286315918,
|
| 841 |
+
"learning_rate": 4.900750952928166e-05,
|
| 842 |
+
"loss": 0.05913209915161133,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.3181897941960444,
|
| 847 |
+
"grad_norm": 0.24644255638122559,
|
| 848 |
+
"learning_rate": 4.8990702359682184e-05,
|
| 849 |
+
"loss": 0.06815173625946044,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.32086365801281785,
|
| 854 |
+
"grad_norm": 2.1861305236816406,
|
| 855 |
+
"learning_rate": 4.897375700434972e-05,
|
| 856 |
+
"loss": 0.04142785966396332,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.32353752182959133,
|
| 861 |
+
"grad_norm": 2.6643004417419434,
|
| 862 |
+
"learning_rate": 4.8956673560889013e-05,
|
| 863 |
+
"loss": 0.05177200436592102,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.3262113856463648,
|
| 868 |
+
"grad_norm": 2.588113784790039,
|
| 869 |
+
"learning_rate": 4.8939452127700195e-05,
|
| 870 |
+
"loss": 0.05783546566963196,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.3288852494631383,
|
| 875 |
+
"grad_norm": 2.419644594192505,
|
| 876 |
+
"learning_rate": 4.8922092803978203e-05,
|
| 877 |
+
"loss": 0.08906854391098022,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.3315591132799118,
|
| 882 |
+
"grad_norm": 0.16949939727783203,
|
| 883 |
+
"learning_rate": 4.890459568971223e-05,
|
| 884 |
+
"loss": 0.10305211544036866,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.33423297709668526,
|
| 889 |
+
"grad_norm": 0.10032984614372253,
|
| 890 |
+
"learning_rate": 4.8886960885685126e-05,
|
| 891 |
+
"loss": 0.06348527669906616,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.33690684091345874,
|
| 896 |
+
"grad_norm": 3.3658738136291504,
|
| 897 |
+
"learning_rate": 4.8869188493472854e-05,
|
| 898 |
+
"loss": 0.06826075911521912,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.3395807047302322,
|
| 903 |
+
"grad_norm": 0.8656186461448669,
|
| 904 |
+
"learning_rate": 4.885127861544386e-05,
|
| 905 |
+
"loss": 0.05929765701293945,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.3422545685470057,
|
| 910 |
+
"grad_norm": 0.1492065042257309,
|
| 911 |
+
"learning_rate": 4.8833231354758496e-05,
|
| 912 |
+
"loss": 0.09429731965065002,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.3449284323637792,
|
| 917 |
+
"grad_norm": 0.6010928153991699,
|
| 918 |
+
"learning_rate": 4.881504681536846e-05,
|
| 919 |
+
"loss": 0.06262240409851075,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.34760229618055266,
|
| 924 |
+
"grad_norm": 1.6506450176239014,
|
| 925 |
+
"learning_rate": 4.879672510201616e-05,
|
| 926 |
+
"loss": 0.061688083410263064,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.35027615999732614,
|
| 931 |
+
"grad_norm": 0.2703142464160919,
|
| 932 |
+
"learning_rate": 4.877826632023412e-05,
|
| 933 |
+
"loss": 0.06175137162208557,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.3529500238140996,
|
| 938 |
+
"grad_norm": 3.1056365966796875,
|
| 939 |
+
"learning_rate": 4.875967057634437e-05,
|
| 940 |
+
"loss": 0.07828506827354431,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.3556238876308731,
|
| 945 |
+
"grad_norm": 0.28790283203125,
|
| 946 |
+
"learning_rate": 4.874093797745784e-05,
|
| 947 |
+
"loss": 0.11355981826782227,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.3582977514476466,
|
| 952 |
+
"grad_norm": 2.3372068405151367,
|
| 953 |
+
"learning_rate": 4.8722068631473746e-05,
|
| 954 |
+
"loss": 0.048267871141433716,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.36097161526442006,
|
| 959 |
+
"grad_norm": 0.12767371535301208,
|
| 960 |
+
"learning_rate": 4.8703062647078976e-05,
|
| 961 |
+
"loss": 0.04319801032543182,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.36364547908119355,
|
| 966 |
+
"grad_norm": 0.5145738124847412,
|
| 967 |
+
"learning_rate": 4.868392013374741e-05,
|
| 968 |
+
"loss": 0.0773090660572052,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.366319342897967,
|
| 973 |
+
"grad_norm": 0.8518500328063965,
|
| 974 |
+
"learning_rate": 4.866464120173937e-05,
|
| 975 |
+
"loss": 0.05149460434913635,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.3689932067147405,
|
| 980 |
+
"grad_norm": 3.6726584434509277,
|
| 981 |
+
"learning_rate": 4.8645225962100924e-05,
|
| 982 |
+
"loss": 0.06896821856498718,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.371667070531514,
|
| 987 |
+
"grad_norm": 1.5626497268676758,
|
| 988 |
+
"learning_rate": 4.862567452666329e-05,
|
| 989 |
+
"loss": 0.047730174660682675,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.37434093434828747,
|
| 994 |
+
"grad_norm": 6.562028884887695,
|
| 995 |
+
"learning_rate": 4.8605987008042144e-05,
|
| 996 |
+
"loss": 0.07060698866844177,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.37701479816506095,
|
| 1001 |
+
"grad_norm": 0.7631726861000061,
|
| 1002 |
+
"learning_rate": 4.8586163519637005e-05,
|
| 1003 |
+
"loss": 0.04944324493408203,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.37968866198183443,
|
| 1008 |
+
"grad_norm": 1.6982293128967285,
|
| 1009 |
+
"learning_rate": 4.8566204175630595e-05,
|
| 1010 |
+
"loss": 0.03000348210334778,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.3823625257986079,
|
| 1015 |
+
"grad_norm": 0.6487429141998291,
|
| 1016 |
+
"learning_rate": 4.854610909098812e-05,
|
| 1017 |
+
"loss": 0.06691416501998901,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.3850363896153814,
|
| 1022 |
+
"grad_norm": 0.7648892402648926,
|
| 1023 |
+
"learning_rate": 4.852587838145668e-05,
|
| 1024 |
+
"loss": 0.05529783964157105,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.38771025343215487,
|
| 1029 |
+
"grad_norm": 0.11601298302412033,
|
| 1030 |
+
"learning_rate": 4.850551216356457e-05,
|
| 1031 |
+
"loss": 0.07780832052230835,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.39038411724892835,
|
| 1036 |
+
"grad_norm": 0.9443137645721436,
|
| 1037 |
+
"learning_rate": 4.8485010554620594e-05,
|
| 1038 |
+
"loss": 0.08007023930549621,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.39305798106570183,
|
| 1043 |
+
"grad_norm": 0.8828252553939819,
|
| 1044 |
+
"learning_rate": 4.846437367271341e-05,
|
| 1045 |
+
"loss": 0.03541453182697296,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.3957318448824753,
|
| 1050 |
+
"grad_norm": 0.21668888628482819,
|
| 1051 |
+
"learning_rate": 4.844360163671083e-05,
|
| 1052 |
+
"loss": 0.08354364633560181,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.3984057086992488,
|
| 1057 |
+
"grad_norm": 0.6840483546257019,
|
| 1058 |
+
"learning_rate": 4.8422694566259194e-05,
|
| 1059 |
+
"loss": 0.045807772874832155,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.4010795725160223,
|
| 1064 |
+
"grad_norm": 1.2754698991775513,
|
| 1065 |
+
"learning_rate": 4.8401652581782584e-05,
|
| 1066 |
+
"loss": 0.053487342596054074,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.40375343633279576,
|
| 1071 |
+
"grad_norm": 0.19012756645679474,
|
| 1072 |
+
"learning_rate": 4.838047580448222e-05,
|
| 1073 |
+
"loss": 0.05881953239440918,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.40642730014956924,
|
| 1078 |
+
"grad_norm": 2.1057698726654053,
|
| 1079 |
+
"learning_rate": 4.835916435633569e-05,
|
| 1080 |
+
"loss": 0.031065690517425536,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.4091011639663427,
|
| 1085 |
+
"grad_norm": 4.188559055328369,
|
| 1086 |
+
"learning_rate": 4.833771836009633e-05,
|
| 1087 |
+
"loss": 0.07205432653427124,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.4117750277831162,
|
| 1092 |
+
"grad_norm": 6.975829124450684,
|
| 1093 |
+
"learning_rate": 4.831613793929242e-05,
|
| 1094 |
+
"loss": 0.04953635036945343,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.4144488915998897,
|
| 1099 |
+
"grad_norm": 4.725269317626953,
|
| 1100 |
+
"learning_rate": 4.8294423218226546e-05,
|
| 1101 |
+
"loss": 0.05965519547462463,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.41712275541666316,
|
| 1106 |
+
"grad_norm": 1.7124755382537842,
|
| 1107 |
+
"learning_rate": 4.827257432197486e-05,
|
| 1108 |
+
"loss": 0.039625433087348935,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.41979661923343664,
|
| 1113 |
+
"grad_norm": 2.6687324047088623,
|
| 1114 |
+
"learning_rate": 4.825059137638636e-05,
|
| 1115 |
+
"loss": 0.05020809769630432,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.4224704830502101,
|
| 1120 |
+
"grad_norm": 1.111640214920044,
|
| 1121 |
+
"learning_rate": 4.822847450808215e-05,
|
| 1122 |
+
"loss": 0.04404452443122864,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.42514434686698366,
|
| 1127 |
+
"grad_norm": 0.2128070890903473,
|
| 1128 |
+
"learning_rate": 4.8206223844454744e-05,
|
| 1129 |
+
"loss": 0.08283355236053466,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.42781821068375714,
|
| 1134 |
+
"grad_norm": 0.10757248103618622,
|
| 1135 |
+
"learning_rate": 4.818383951366729e-05,
|
| 1136 |
+
"loss": 0.08568671345710754,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.4304920745005306,
|
| 1141 |
+
"grad_norm": 0.08344592899084091,
|
| 1142 |
+
"learning_rate": 4.816132164465289e-05,
|
| 1143 |
+
"loss": 0.0426956832408905,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.4331659383173041,
|
| 1148 |
+
"grad_norm": 0.5657751560211182,
|
| 1149 |
+
"learning_rate": 4.813867036711378e-05,
|
| 1150 |
+
"loss": 0.04971776902675629,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.4358398021340776,
|
| 1155 |
+
"grad_norm": 2.1529288291931152,
|
| 1156 |
+
"learning_rate": 4.8115885811520654e-05,
|
| 1157 |
+
"loss": 0.025386181473731995,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.43851366595085106,
|
| 1162 |
+
"grad_norm": 4.228519916534424,
|
| 1163 |
+
"learning_rate": 4.809296810911188e-05,
|
| 1164 |
+
"loss": 0.06401395201683044,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.44118752976762454,
|
| 1169 |
+
"grad_norm": 6.770420551300049,
|
| 1170 |
+
"learning_rate": 4.806991739189274e-05,
|
| 1171 |
+
"loss": 0.16425553560256959,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.443861393584398,
|
| 1176 |
+
"grad_norm": 0.5303187370300293,
|
| 1177 |
+
"learning_rate": 4.804673379263467e-05,
|
| 1178 |
+
"loss": 0.045900467038154605,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.4465352574011715,
|
| 1183 |
+
"grad_norm": 0.221473827958107,
|
| 1184 |
+
"learning_rate": 4.802341744487453e-05,
|
| 1185 |
+
"loss": 0.07529735565185547,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.449209121217945,
|
| 1190 |
+
"grad_norm": 3.48736834526062,
|
| 1191 |
+
"learning_rate": 4.799996848291378e-05,
|
| 1192 |
+
"loss": 0.062433135509490964,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.45188298503471847,
|
| 1197 |
+
"grad_norm": 2.650038242340088,
|
| 1198 |
+
"learning_rate": 4.797638704181774e-05,
|
| 1199 |
+
"loss": 0.03762982189655304,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.45455684885149195,
|
| 1204 |
+
"grad_norm": 3.159665584564209,
|
| 1205 |
+
"learning_rate": 4.795267325741483e-05,
|
| 1206 |
+
"loss": 0.04745924174785614,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.4572307126682654,
|
| 1211 |
+
"grad_norm": 0.8763885498046875,
|
| 1212 |
+
"learning_rate": 4.7928827266295715e-05,
|
| 1213 |
+
"loss": 0.07380253076553345,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.4599045764850389,
|
| 1218 |
+
"grad_norm": 0.1779366433620453,
|
| 1219 |
+
"learning_rate": 4.790484920581262e-05,
|
| 1220 |
+
"loss": 0.045916372537612916,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.4625784403018124,
|
| 1225 |
+
"grad_norm": 1.1228729486465454,
|
| 1226 |
+
"learning_rate": 4.7880739214078454e-05,
|
| 1227 |
+
"loss": 0.04461723566055298,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.46525230411858587,
|
| 1232 |
+
"grad_norm": 0.1629919707775116,
|
| 1233 |
+
"learning_rate": 4.785649742996605e-05,
|
| 1234 |
+
"loss": 0.017159442603588104,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.46792616793535935,
|
| 1239 |
+
"grad_norm": 3.583951473236084,
|
| 1240 |
+
"learning_rate": 4.783212399310737e-05,
|
| 1241 |
+
"loss": 0.047145146131515506,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.47060003175213283,
|
| 1246 |
+
"grad_norm": 0.9766237139701843,
|
| 1247 |
+
"learning_rate": 4.780761904389267e-05,
|
| 1248 |
+
"loss": 0.050229442119598386,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.4732738955689063,
|
| 1253 |
+
"grad_norm": 0.05617872253060341,
|
| 1254 |
+
"learning_rate": 4.778298272346976e-05,
|
| 1255 |
+
"loss": 0.024862812459468843,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.4759477593856798,
|
| 1260 |
+
"grad_norm": 1.3586453199386597,
|
| 1261 |
+
"learning_rate": 4.775821517374308e-05,
|
| 1262 |
+
"loss": 0.02117772251367569,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.4786216232024533,
|
| 1267 |
+
"grad_norm": 1.2116742134094238,
|
| 1268 |
+
"learning_rate": 4.7733316537373006e-05,
|
| 1269 |
+
"loss": 0.03060794174671173,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.48129548701922675,
|
| 1274 |
+
"grad_norm": 0.39403238892555237,
|
| 1275 |
+
"learning_rate": 4.770828695777493e-05,
|
| 1276 |
+
"loss": 0.05482668280601501,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.48396935083600023,
|
| 1281 |
+
"grad_norm": 0.9248486161231995,
|
| 1282 |
+
"learning_rate": 4.7683126579118495e-05,
|
| 1283 |
+
"loss": 0.03612814247608185,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.4866432146527737,
|
| 1288 |
+
"grad_norm": 0.1624649167060852,
|
| 1289 |
+
"learning_rate": 4.7657835546326736e-05,
|
| 1290 |
+
"loss": 0.04334873259067536,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.4893170784695472,
|
| 1295 |
+
"grad_norm": 0.5321119427680969,
|
| 1296 |
+
"learning_rate": 4.763241400507524e-05,
|
| 1297 |
+
"loss": 0.0461233913898468,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.4919909422863207,
|
| 1302 |
+
"grad_norm": 0.34861093759536743,
|
| 1303 |
+
"learning_rate": 4.760686210179133e-05,
|
| 1304 |
+
"loss": 0.024829554557800292,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.49466480610309416,
|
| 1309 |
+
"grad_norm": 1.2561241388320923,
|
| 1310 |
+
"learning_rate": 4.758117998365322e-05,
|
| 1311 |
+
"loss": 0.03157005608081818,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.49733866991986764,
|
| 1316 |
+
"grad_norm": 0.8691341280937195,
|
| 1317 |
+
"learning_rate": 4.7555367798589146e-05,
|
| 1318 |
+
"loss": 0.04310203492641449,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.5000125337366411,
|
| 1323 |
+
"grad_norm": 0.3134572505950928,
|
| 1324 |
+
"learning_rate": 4.752942569527653e-05,
|
| 1325 |
+
"loss": 0.03796039223670959,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.5026863975534146,
|
| 1330 |
+
"grad_norm": 2.3359289169311523,
|
| 1331 |
+
"learning_rate": 4.75033538231411e-05,
|
| 1332 |
+
"loss": 0.055599170923233035,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.5053602613701881,
|
| 1337 |
+
"grad_norm": 7.426175594329834,
|
| 1338 |
+
"learning_rate": 4.747715233235608e-05,
|
| 1339 |
+
"loss": 0.054436272382736205,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.5080341251869616,
|
| 1344 |
+
"grad_norm": 0.5940203070640564,
|
| 1345 |
+
"learning_rate": 4.745082137384128e-05,
|
| 1346 |
+
"loss": 0.03682814538478851,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.510707989003735,
|
| 1351 |
+
"grad_norm": 0.22821389138698578,
|
| 1352 |
+
"learning_rate": 4.7424361099262225e-05,
|
| 1353 |
+
"loss": 0.051123309135437014,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.5133818528205085,
|
| 1358 |
+
"grad_norm": 8.20633602142334,
|
| 1359 |
+
"learning_rate": 4.739777166102932e-05,
|
| 1360 |
+
"loss": 0.0704378604888916,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.516055716637282,
|
| 1365 |
+
"grad_norm": 3.023848533630371,
|
| 1366 |
+
"learning_rate": 4.737105321229694e-05,
|
| 1367 |
+
"loss": 0.03368058800697327,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.5187295804540555,
|
| 1372 |
+
"grad_norm": 0.07666649669408798,
|
| 1373 |
+
"learning_rate": 4.7344205906962555e-05,
|
| 1374 |
+
"loss": 0.03665303289890289,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.521403444270829,
|
| 1379 |
+
"grad_norm": 0.7571629881858826,
|
| 1380 |
+
"learning_rate": 4.731722989966585e-05,
|
| 1381 |
+
"loss": 0.058415502309799194,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.5240773080876024,
|
| 1386 |
+
"grad_norm": 3.2599120140075684,
|
| 1387 |
+
"learning_rate": 4.7290125345787816e-05,
|
| 1388 |
+
"loss": 0.07323018908500671,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.5267511719043759,
|
| 1393 |
+
"grad_norm": 0.28930988907814026,
|
| 1394 |
+
"learning_rate": 4.7262892401449886e-05,
|
| 1395 |
+
"loss": 0.054371267557144165,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.5294250357211494,
|
| 1400 |
+
"grad_norm": 2.2296454906463623,
|
| 1401 |
+
"learning_rate": 4.7235531223513004e-05,
|
| 1402 |
+
"loss": 0.040819621086120604,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.5320988995379229,
|
| 1407 |
+
"grad_norm": 0.11608211696147919,
|
| 1408 |
+
"learning_rate": 4.720804196957675e-05,
|
| 1409 |
+
"loss": 0.05215579271316528,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.5347727633546964,
|
| 1414 |
+
"grad_norm": 1.1587547063827515,
|
| 1415 |
+
"learning_rate": 4.7180424797978415e-05,
|
| 1416 |
+
"loss": 0.026277875900268553,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.5374466271714698,
|
| 1421 |
+
"grad_norm": 0.06253435462713242,
|
| 1422 |
+
"learning_rate": 4.7152679867792074e-05,
|
| 1423 |
+
"loss": 0.02574407756328583,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.5401204909882433,
|
| 1428 |
+
"grad_norm": 1.3441458940505981,
|
| 1429 |
+
"learning_rate": 4.71248073388277e-05,
|
| 1430 |
+
"loss": 0.05538107752799988,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.5427943548050168,
|
| 1435 |
+
"grad_norm": 0.48076340556144714,
|
| 1436 |
+
"learning_rate": 4.7096807371630236e-05,
|
| 1437 |
+
"loss": 0.047986540198326114,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.5454682186217903,
|
| 1442 |
+
"grad_norm": 0.5924936532974243,
|
| 1443 |
+
"learning_rate": 4.706868012747867e-05,
|
| 1444 |
+
"loss": 0.05463914275169372,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.7673980604481948,
|
| 1449 |
+
"grad_norm": 0.4317747950553894,
|
| 1450 |
+
"learning_rate": 4.431151627307268e-05,
|
| 1451 |
+
"loss": 0.020150287946065267,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.7711414656211129,
|
| 1456 |
+
"grad_norm": 0.7817056179046631,
|
| 1457 |
+
"learning_rate": 4.425806509248848e-05,
|
| 1458 |
+
"loss": 0.029157504439353943,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.7748848707940308,
|
| 1463 |
+
"grad_norm": 4.8055219650268555,
|
| 1464 |
+
"learning_rate": 4.420439652052499e-05,
|
| 1465 |
+
"loss": 0.05370526909828186,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.7786282759669488,
|
| 1470 |
+
"grad_norm": 0.5026630759239197,
|
| 1471 |
+
"learning_rate": 4.415051116301072e-05,
|
| 1472 |
+
"loss": 0.03046708106994629,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.7823716811398669,
|
| 1477 |
+
"grad_norm": 0.10916770249605179,
|
| 1478 |
+
"learning_rate": 4.409640962822132e-05,
|
| 1479 |
+
"loss": 0.024332405626773836,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.7861150863127849,
|
| 1484 |
+
"grad_norm": 0.559330403804779,
|
| 1485 |
+
"learning_rate": 4.404209252687275e-05,
|
| 1486 |
+
"loss": 0.016460689902305602,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.789858491485703,
|
| 1491 |
+
"grad_norm": 1.0136655569076538,
|
| 1492 |
+
"learning_rate": 4.398756047211431e-05,
|
| 1493 |
+
"loss": 0.03609788417816162,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.7936018966586209,
|
| 1498 |
+
"grad_norm": 1.597655177116394,
|
| 1499 |
+
"learning_rate": 4.39328140795218e-05,
|
| 1500 |
+
"loss": 0.03543405830860138,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.797345301831539,
|
| 1505 |
+
"grad_norm": 0.21154026687145233,
|
| 1506 |
+
"learning_rate": 4.387785396709052e-05,
|
| 1507 |
+
"loss": 0.06029776334762573,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.801088707004457,
|
| 1512 |
+
"grad_norm": 1.3103065490722656,
|
| 1513 |
+
"learning_rate": 4.382268075522831e-05,
|
| 1514 |
+
"loss": 0.05260368585586548,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.804832112177375,
|
| 1519 |
+
"grad_norm": 0.07966366410255432,
|
| 1520 |
+
"learning_rate": 4.3767295066748564e-05,
|
| 1521 |
+
"loss": 0.02691380977630615,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.8085755173502931,
|
| 1526 |
+
"grad_norm": 0.17526337504386902,
|
| 1527 |
+
"learning_rate": 4.371169752686316e-05,
|
| 1528 |
+
"loss": 0.02926369905471802,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.812318922523211,
|
| 1533 |
+
"grad_norm": 1.6024236679077148,
|
| 1534 |
+
"learning_rate": 4.3655888763175436e-05,
|
| 1535 |
+
"loss": 0.09227399826049805,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.8160623276961291,
|
| 1540 |
+
"grad_norm": 2.4870424270629883,
|
| 1541 |
+
"learning_rate": 4.3599869405673085e-05,
|
| 1542 |
+
"loss": 0.05842729210853577,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.8198057328690471,
|
| 1547 |
+
"grad_norm": 0.27151742577552795,
|
| 1548 |
+
"learning_rate": 4.354364008672106e-05,
|
| 1549 |
+
"loss": 0.03142921030521393,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.8235491380419652,
|
| 1554 |
+
"grad_norm": 0.06985723227262497,
|
| 1555 |
+
"learning_rate": 4.3487201441054435e-05,
|
| 1556 |
+
"loss": 0.021850427985191344,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.8272925432148831,
|
| 1561 |
+
"grad_norm": 0.9921057224273682,
|
| 1562 |
+
"learning_rate": 4.343055410577122e-05,
|
| 1563 |
+
"loss": 0.03694712221622467,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.8310359483878011,
|
| 1568 |
+
"grad_norm": 3.330061912536621,
|
| 1569 |
+
"learning_rate": 4.3373698720325176e-05,
|
| 1570 |
+
"loss": 0.04195646345615387,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.8347793535607192,
|
| 1575 |
+
"grad_norm": 4.173893928527832,
|
| 1576 |
+
"learning_rate": 4.331663592651862e-05,
|
| 1577 |
+
"loss": 0.025657615065574645,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.8385227587336372,
|
| 1582 |
+
"grad_norm": 0.514607846736908,
|
| 1583 |
+
"learning_rate": 4.3259366368495167e-05,
|
| 1584 |
+
"loss": 0.018329595029354096,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.8422661639065553,
|
| 1589 |
+
"grad_norm": 0.7693771719932556,
|
| 1590 |
+
"learning_rate": 4.320189069273243e-05,
|
| 1591 |
+
"loss": 0.0356669157743454,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.8460095690794732,
|
| 1596 |
+
"grad_norm": 0.27708783745765686,
|
| 1597 |
+
"learning_rate": 4.3144209548034766e-05,
|
| 1598 |
+
"loss": 0.01452532410621643,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.8497529742523913,
|
| 1603 |
+
"grad_norm": 2.8631412982940674,
|
| 1604 |
+
"learning_rate": 4.3086323585525915e-05,
|
| 1605 |
+
"loss": 0.020461027324199677,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.8534963794253093,
|
| 1610 |
+
"grad_norm": 0.09945106506347656,
|
| 1611 |
+
"learning_rate": 4.3028233458641696e-05,
|
| 1612 |
+
"loss": 0.06334148645401001,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.8572397845982274,
|
| 1617 |
+
"grad_norm": 0.43821537494659424,
|
| 1618 |
+
"learning_rate": 4.2969939823122586e-05,
|
| 1619 |
+
"loss": 0.024606382846832274,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.8609831897711454,
|
| 1624 |
+
"grad_norm": 0.04385611414909363,
|
| 1625 |
+
"learning_rate": 4.291144333700633e-05,
|
| 1626 |
+
"loss": 0.047047534584999086,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.8647265949440633,
|
| 1631 |
+
"grad_norm": 0.9393408298492432,
|
| 1632 |
+
"learning_rate": 4.2852744660620515e-05,
|
| 1633 |
+
"loss": 0.05728556513786316,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.8684700001169814,
|
| 1638 |
+
"grad_norm": 0.07993622869253159,
|
| 1639 |
+
"learning_rate": 4.279384445657514e-05,
|
| 1640 |
+
"loss": 0.04038055539131165,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.8722134052898994,
|
| 1645 |
+
"grad_norm": 1.3048001527786255,
|
| 1646 |
+
"learning_rate": 4.2734743389755096e-05,
|
| 1647 |
+
"loss": 0.04012865424156189,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.8759568104628175,
|
| 1652 |
+
"grad_norm": 1.23117196559906,
|
| 1653 |
+
"learning_rate": 4.267544212731268e-05,
|
| 1654 |
+
"loss": 0.02873348295688629,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.8797002156357355,
|
| 1659 |
+
"grad_norm": 0.058857761323451996,
|
| 1660 |
+
"learning_rate": 4.261594133866007e-05,
|
| 1661 |
+
"loss": 0.017506703734397888,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.8834436208086535,
|
| 1666 |
+
"grad_norm": 0.061513226479291916,
|
| 1667 |
+
"learning_rate": 4.255624169546175e-05,
|
| 1668 |
+
"loss": 0.04879063069820404,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.8871870259815715,
|
| 1673 |
+
"grad_norm": 1.559389591217041,
|
| 1674 |
+
"learning_rate": 4.249634387162696e-05,
|
| 1675 |
+
"loss": 0.05136169791221619,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.8909304311544896,
|
| 1680 |
+
"grad_norm": 0.42972251772880554,
|
| 1681 |
+
"learning_rate": 4.243624854330206e-05,
|
| 1682 |
+
"loss": 0.033595073223114016,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.8946738363274076,
|
| 1687 |
+
"grad_norm": 2.1139724254608154,
|
| 1688 |
+
"learning_rate": 4.237595638886288e-05,
|
| 1689 |
+
"loss": 0.0396845281124115,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.8984172415003256,
|
| 1694 |
+
"grad_norm": 10.029800415039062,
|
| 1695 |
+
"learning_rate": 4.231546808890713e-05,
|
| 1696 |
+
"loss": 0.02165188193321228,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.9021606466732436,
|
| 1701 |
+
"grad_norm": 0.07373173534870148,
|
| 1702 |
+
"learning_rate": 4.225478432624665e-05,
|
| 1703 |
+
"loss": 0.011127500236034394,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.9059040518461616,
|
| 1708 |
+
"grad_norm": 0.04600556939840317,
|
| 1709 |
+
"learning_rate": 4.219390578589973e-05,
|
| 1710 |
+
"loss": 0.02061241865158081,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.9096474570190797,
|
| 1715 |
+
"grad_norm": 2.570568799972534,
|
| 1716 |
+
"learning_rate": 4.213283315508337e-05,
|
| 1717 |
+
"loss": 0.028173595666885376,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.9133908621919977,
|
| 1722 |
+
"grad_norm": 1.7165521383285522,
|
| 1723 |
+
"learning_rate": 4.207156712320555e-05,
|
| 1724 |
+
"loss": 0.03346221446990967,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.9171342673649158,
|
| 1729 |
+
"grad_norm": 0.20669779181480408,
|
| 1730 |
+
"learning_rate": 4.20101083818574e-05,
|
| 1731 |
+
"loss": 0.03200317621231079,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.9208776725378337,
|
| 1736 |
+
"grad_norm": 0.10834258794784546,
|
| 1737 |
+
"learning_rate": 4.194845762480544e-05,
|
| 1738 |
+
"loss": 0.022559843957424164,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.9246210777107517,
|
| 1743 |
+
"grad_norm": 0.11646590381860733,
|
| 1744 |
+
"learning_rate": 4.188661554798369e-05,
|
| 1745 |
+
"loss": 0.02637559175491333,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.9283644828836698,
|
| 1750 |
+
"grad_norm": 0.8932350277900696,
|
| 1751 |
+
"learning_rate": 4.1824582849485884e-05,
|
| 1752 |
+
"loss": 0.03692157864570618,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.9321078880565878,
|
| 1757 |
+
"grad_norm": 0.1985676884651184,
|
| 1758 |
+
"learning_rate": 4.176236022955755e-05,
|
| 1759 |
+
"loss": 0.028895530104637145,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.9358512932295058,
|
| 1764 |
+
"grad_norm": 0.9143677353858948,
|
| 1765 |
+
"learning_rate": 4.16999483905881e-05,
|
| 1766 |
+
"loss": 0.030196067690849305,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.9395946984024238,
|
| 1771 |
+
"grad_norm": 0.4041534662246704,
|
| 1772 |
+
"learning_rate": 4.163734803710294e-05,
|
| 1773 |
+
"loss": 0.02996099293231964,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.9433381035753419,
|
| 1778 |
+
"grad_norm": 0.11670702695846558,
|
| 1779 |
+
"learning_rate": 4.157455987575545e-05,
|
| 1780 |
+
"loss": 0.030414831638336182,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.9470815087482599,
|
| 1785 |
+
"grad_norm": 1.33324134349823,
|
| 1786 |
+
"learning_rate": 4.1511584615319075e-05,
|
| 1787 |
+
"loss": 0.024954181909561158,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.950824913921178,
|
| 1792 |
+
"grad_norm": 0.366225004196167,
|
| 1793 |
+
"learning_rate": 4.144842296667929e-05,
|
| 1794 |
+
"loss": 0.020582753419876098,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.9545683190940959,
|
| 1799 |
+
"grad_norm": 2.255720615386963,
|
| 1800 |
+
"learning_rate": 4.138507564282558e-05,
|
| 1801 |
+
"loss": 0.047728517651557924,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.9583117242670139,
|
| 1806 |
+
"grad_norm": 0.4103017747402191,
|
| 1807 |
+
"learning_rate": 4.1321543358843385e-05,
|
| 1808 |
+
"loss": 0.021700438857078553,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.962055129439932,
|
| 1813 |
+
"grad_norm": 1.1296149492263794,
|
| 1814 |
+
"learning_rate": 4.125782683190606e-05,
|
| 1815 |
+
"loss": 0.013979548215866089,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.96579853461285,
|
| 1820 |
+
"grad_norm": 0.09410075098276138,
|
| 1821 |
+
"learning_rate": 4.119392678126673e-05,
|
| 1822 |
+
"loss": 0.04596070945262909,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.9695419397857681,
|
| 1827 |
+
"grad_norm": 8.157197952270508,
|
| 1828 |
+
"learning_rate": 4.11298439282502e-05,
|
| 1829 |
+
"loss": 0.029448685050010682,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.973285344958686,
|
| 1834 |
+
"grad_norm": 0.04135945811867714,
|
| 1835 |
+
"learning_rate": 4.106557899624482e-05,
|
| 1836 |
+
"loss": 0.03777257204055786,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.9770287501316041,
|
| 1841 |
+
"grad_norm": 0.05247588828206062,
|
| 1842 |
+
"learning_rate": 4.1001132710694304e-05,
|
| 1843 |
+
"loss": 0.04280385076999664,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.9807721553045221,
|
| 1848 |
+
"grad_norm": 2.1289517879486084,
|
| 1849 |
+
"learning_rate": 4.093650579908953e-05,
|
| 1850 |
+
"loss": 0.015029631555080414,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.9845155604774402,
|
| 1855 |
+
"grad_norm": 0.08014921844005585,
|
| 1856 |
+
"learning_rate": 4.087169899096037e-05,
|
| 1857 |
+
"loss": 0.02902546226978302,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.9882589656503582,
|
| 1862 |
+
"grad_norm": 0.7032691836357117,
|
| 1863 |
+
"learning_rate": 4.080671301786741e-05,
|
| 1864 |
+
"loss": 0.03703016638755798,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.9920023708232761,
|
| 1869 |
+
"grad_norm": 0.02973131090402603,
|
| 1870 |
+
"learning_rate": 4.0741548613393675e-05,
|
| 1871 |
+
"loss": 0.027672907710075377,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.9957457759961942,
|
| 1876 |
+
"grad_norm": 4.2270379066467285,
|
| 1877 |
+
"learning_rate": 4.067620651313647e-05,
|
| 1878 |
+
"loss": 0.06081045866012573,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.9994891811691122,
|
| 1883 |
+
"grad_norm": 0.4717981815338135,
|
| 1884 |
+
"learning_rate": 4.0610687454698906e-05,
|
| 1885 |
+
"loss": 0.009954025596380233,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.0031818943969804,
|
| 1890 |
+
"grad_norm": 1.7803936004638672,
|
| 1891 |
+
"learning_rate": 4.0544992177681685e-05,
|
| 1892 |
+
"loss": 0.016653330624103548,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.0069252995698983,
|
| 1897 |
+
"grad_norm": 0.46166855096817017,
|
| 1898 |
+
"learning_rate": 4.047912142367473e-05,
|
| 1899 |
+
"loss": 0.013706094026565552,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.0106687047428164,
|
| 1904 |
+
"grad_norm": 5.546389102935791,
|
| 1905 |
+
"learning_rate": 4.04130759362488e-05,
|
| 1906 |
+
"loss": 0.031427782773971555,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.0144121099157344,
|
| 1911 |
+
"grad_norm": 0.2795480787754059,
|
| 1912 |
+
"learning_rate": 4.034685646094711e-05,
|
| 1913 |
+
"loss": 0.006191424652934074,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.0181555150886523,
|
| 1918 |
+
"grad_norm": 1.0172348022460938,
|
| 1919 |
+
"learning_rate": 4.028046374527689e-05,
|
| 1920 |
+
"loss": 0.01738315522670746,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.0218989202615705,
|
| 1925 |
+
"grad_norm": 2.8326938152313232,
|
| 1926 |
+
"learning_rate": 4.021389853870095e-05,
|
| 1927 |
+
"loss": 0.02997104823589325,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.0256423254344884,
|
| 1932 |
+
"grad_norm": 0.03412143886089325,
|
| 1933 |
+
"learning_rate": 4.0147161592629306e-05,
|
| 1934 |
+
"loss": 0.04240097105503082,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.0293857306074066,
|
| 1939 |
+
"grad_norm": 2.6316885948181152,
|
| 1940 |
+
"learning_rate": 4.008025366041055e-05,
|
| 1941 |
+
"loss": 0.008842331916093826,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.0331291357803245,
|
| 1946 |
+
"grad_norm": 0.8536480069160461,
|
| 1947 |
+
"learning_rate": 4.001317549732345e-05,
|
| 1948 |
+
"loss": 0.013698439300060272,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.0368725409532424,
|
| 1953 |
+
"grad_norm": 0.029558265581727028,
|
| 1954 |
+
"learning_rate": 3.99459278605684e-05,
|
| 1955 |
+
"loss": 0.029040488600730895,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.0406159461261606,
|
| 1960 |
+
"grad_norm": 0.03936990723013878,
|
| 1961 |
+
"learning_rate": 3.9878511509258866e-05,
|
| 1962 |
+
"loss": 0.014562596380710603,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.0443593512990785,
|
| 1967 |
+
"grad_norm": 0.5267767310142517,
|
| 1968 |
+
"learning_rate": 3.9810927204412803e-05,
|
| 1969 |
+
"loss": 0.008023031055927277,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.0481027564719967,
|
| 1974 |
+
"grad_norm": 0.2755996584892273,
|
| 1975 |
+
"learning_rate": 3.974317570894413e-05,
|
| 1976 |
+
"loss": 0.01961633116006851,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.0518461616449146,
|
| 1981 |
+
"grad_norm": 1.5329219102859497,
|
| 1982 |
+
"learning_rate": 3.9675257787654e-05,
|
| 1983 |
+
"loss": 0.016856445372104643,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.0555895668178326,
|
| 1988 |
+
"grad_norm": 0.06532379984855652,
|
| 1989 |
+
"learning_rate": 3.960717420722227e-05,
|
| 1990 |
+
"loss": 0.016211540997028352,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.0593329719907507,
|
| 1995 |
+
"grad_norm": 0.2080603390932083,
|
| 1996 |
+
"learning_rate": 3.953892573619883e-05,
|
| 1997 |
+
"loss": 0.013586725294589996,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.0630763771636687,
|
| 2002 |
+
"grad_norm": 0.46395453810691833,
|
| 2003 |
+
"learning_rate": 3.947051314499489e-05,
|
| 2004 |
+
"loss": 0.011446774005889893,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.0668197823365868,
|
| 2009 |
+
"grad_norm": 0.5086866617202759,
|
| 2010 |
+
"learning_rate": 3.94019372058743e-05,
|
| 2011 |
+
"loss": 0.012980939447879791,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.0705631875095047,
|
| 2016 |
+
"grad_norm": 0.20874151587486267,
|
| 2017 |
+
"learning_rate": 3.933319869294483e-05,
|
| 2018 |
+
"loss": 0.012278204411268234,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.0743065926824227,
|
| 2023 |
+
"grad_norm": 0.05940713733434677,
|
| 2024 |
+
"learning_rate": 3.9264298382149455e-05,
|
| 2025 |
+
"loss": 0.026001006364822388,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.0780499978553408,
|
| 2030 |
+
"grad_norm": 2.1957216262817383,
|
| 2031 |
+
"learning_rate": 3.919523705125757e-05,
|
| 2032 |
+
"loss": 0.014651447534561157,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 1.0817934030282588,
|
| 2037 |
+
"grad_norm": 0.08526397496461868,
|
| 2038 |
+
"learning_rate": 3.9126015479856205e-05,
|
| 2039 |
+
"loss": 0.018336918950080872,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.085536808201177,
|
| 2044 |
+
"grad_norm": 0.6159129738807678,
|
| 2045 |
+
"learning_rate": 3.9056634449341256e-05,
|
| 2046 |
+
"loss": 0.005850336700677872,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.0892802133740949,
|
| 2051 |
+
"grad_norm": 0.5908898711204529,
|
| 2052 |
+
"learning_rate": 3.898709474290864e-05,
|
| 2053 |
+
"loss": 0.03193281590938568,
|
| 2054 |
+
"step": 5820
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.0930236185470128,
|
| 2058 |
+
"grad_norm": 0.3759937882423401,
|
| 2059 |
+
"learning_rate": 3.8917397145545454e-05,
|
| 2060 |
+
"loss": 0.017000101506710052,
|
| 2061 |
+
"step": 5840
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.096767023719931,
|
| 2065 |
+
"grad_norm": 0.07813868671655655,
|
| 2066 |
+
"learning_rate": 3.884754244402113e-05,
|
| 2067 |
+
"loss": 0.03016698658466339,
|
| 2068 |
+
"step": 5860
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 1.100510428892849,
|
| 2072 |
+
"grad_norm": 0.022068507969379425,
|
| 2073 |
+
"learning_rate": 3.877753142687852e-05,
|
| 2074 |
+
"loss": 0.02659503519535065,
|
| 2075 |
+
"step": 5880
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 1.104253834065767,
|
| 2079 |
+
"grad_norm": 1.6575422286987305,
|
| 2080 |
+
"learning_rate": 3.8707364884425064e-05,
|
| 2081 |
+
"loss": 0.016750173270702363,
|
| 2082 |
+
"step": 5900
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 1.107997239238685,
|
| 2086 |
+
"grad_norm": 1.673715591430664,
|
| 2087 |
+
"learning_rate": 3.863704360872378e-05,
|
| 2088 |
+
"loss": 0.03380795419216156,
|
| 2089 |
+
"step": 5920
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 1.111740644411603,
|
| 2093 |
+
"grad_norm": 0.04445308819413185,
|
| 2094 |
+
"learning_rate": 3.8566568393584366e-05,
|
| 2095 |
+
"loss": 0.029713729023933412,
|
| 2096 |
+
"step": 5940
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 1.115484049584521,
|
| 2100 |
+
"grad_norm": 0.15362456440925598,
|
| 2101 |
+
"learning_rate": 3.8495940034554283e-05,
|
| 2102 |
+
"loss": 0.01640743464231491,
|
| 2103 |
+
"step": 5960
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.119227454757439,
|
| 2107 |
+
"grad_norm": 0.04955374822020531,
|
| 2108 |
+
"learning_rate": 3.8425159328909684e-05,
|
| 2109 |
+
"loss": 0.010707255452871323,
|
| 2110 |
+
"step": 5980
|
| 2111 |
+
}
|
| 2112 |
+
],
|
| 2113 |
+
"logging_steps": 20,
|
| 2114 |
+
"max_steps": 18701,
|
| 2115 |
+
"num_input_tokens_seen": 0,
|
| 2116 |
+
"num_train_epochs": 4,
|
| 2117 |
+
"save_steps": 1000000000,
|
| 2118 |
+
"stateful_callbacks": {
|
| 2119 |
+
"TrainerControl": {
|
| 2120 |
+
"args": {
|
| 2121 |
+
"should_epoch_stop": false,
|
| 2122 |
+
"should_evaluate": false,
|
| 2123 |
+
"should_log": false,
|
| 2124 |
+
"should_save": true,
|
| 2125 |
+
"should_training_stop": false
|
| 2126 |
+
},
|
| 2127 |
+
"attributes": {}
|
| 2128 |
+
}
|
| 2129 |
+
},
|
| 2130 |
+
"total_flos": 2.298013872961997e+16,
|
| 2131 |
+
"train_batch_size": 1,
|
| 2132 |
+
"trial_name": null,
|
| 2133 |
+
"trial_params": null
|
| 2134 |
+
}
|