Upload folder using huggingface_hub
Browse files
slots/3/checkpoint-8335/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": 1,
|
| 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/3/checkpoint-8335/trainer_state.json
ADDED
|
@@ -0,0 +1,2953 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.489067427345171,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 8335,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 2.617016627215141e-05,
|
| 14 |
+
"grad_norm": 122.33340454101562,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 419.9996032714844,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0005234033254430282,
|
| 21 |
+
"grad_norm": 90.82410430908203,
|
| 22 |
+
"learning_rate": 4.999999511980259e-05,
|
| 23 |
+
"loss": 428.18197471217104,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0010468066508860564,
|
| 28 |
+
"grad_norm": 59.52641296386719,
|
| 29 |
+
"learning_rate": 4.999997943828395e-05,
|
| 30 |
+
"loss": 366.3947265625,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0015702099763290846,
|
| 35 |
+
"grad_norm": 41.61277389526367,
|
| 36 |
+
"learning_rate": 4.999995294193233e-05,
|
| 37 |
+
"loss": 325.6545166015625,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0020936133017721128,
|
| 42 |
+
"grad_norm": 46.67556381225586,
|
| 43 |
+
"learning_rate": 4.999991563075919e-05,
|
| 44 |
+
"loss": 310.3648193359375,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.002617016627215141,
|
| 49 |
+
"grad_norm": 47.65824890136719,
|
| 50 |
+
"learning_rate": 4.999986750478065e-05,
|
| 51 |
+
"loss": 303.49443359375,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.003140419952658169,
|
| 56 |
+
"grad_norm": 62.748138427734375,
|
| 57 |
+
"learning_rate": 4.999980856401755e-05,
|
| 58 |
+
"loss": 322.389990234375,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.007327646556202395,
|
| 63 |
+
"grad_norm": 48.20169448852539,
|
| 64 |
+
"learning_rate": 4.999895523943923e-05,
|
| 65 |
+
"loss": 288.82497336647725,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.008374453207088451,
|
| 70 |
+
"grad_norm": 68.66603088378906,
|
| 71 |
+
"learning_rate": 4.99986329623214e-05,
|
| 72 |
+
"loss": 293.87509765625,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.009421259857974507,
|
| 77 |
+
"grad_norm": 61.69401931762695,
|
| 78 |
+
"learning_rate": 4.999826742820622e-05,
|
| 79 |
+
"loss": 324.10361328125,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.010468066508860565,
|
| 84 |
+
"grad_norm": 56.778385162353516,
|
| 85 |
+
"learning_rate": 4.999785863772618e-05,
|
| 86 |
+
"loss": 274.875537109375,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.01151487315974662,
|
| 91 |
+
"grad_norm": 71.3816146850586,
|
| 92 |
+
"learning_rate": 4.999740659158865e-05,
|
| 93 |
+
"loss": 281.357080078125,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.012561679810632677,
|
| 98 |
+
"grad_norm": 75.08697509765625,
|
| 99 |
+
"learning_rate": 4.9996911290575844e-05,
|
| 100 |
+
"loss": 263.852783203125,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.013608486461518732,
|
| 105 |
+
"grad_norm": 53.91341781616211,
|
| 106 |
+
"learning_rate": 4.999637273554481e-05,
|
| 107 |
+
"loss": 258.546630859375,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.01465529311240479,
|
| 112 |
+
"grad_norm": 56.39162063598633,
|
| 113 |
+
"learning_rate": 4.9995790927427446e-05,
|
| 114 |
+
"loss": 270.5470703125,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.015702099763290844,
|
| 119 |
+
"grad_norm": 62.93777084350586,
|
| 120 |
+
"learning_rate": 4.99951658672305e-05,
|
| 121 |
+
"loss": 252.8383056640625,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.016748906414176902,
|
| 126 |
+
"grad_norm": 61.51578903198242,
|
| 127 |
+
"learning_rate": 4.9994497556035567e-05,
|
| 128 |
+
"loss": 287.8167724609375,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.01779571306506296,
|
| 133 |
+
"grad_norm": 62.75235366821289,
|
| 134 |
+
"learning_rate": 4.9993785994999074e-05,
|
| 135 |
+
"loss": 258.673779296875,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.018842519715949014,
|
| 140 |
+
"grad_norm": 56.2258186340332,
|
| 141 |
+
"learning_rate": 4.999303118535229e-05,
|
| 142 |
+
"loss": 252.9609375,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.01988932636683507,
|
| 147 |
+
"grad_norm": 54.998130798339844,
|
| 148 |
+
"learning_rate": 4.9992233128401314e-05,
|
| 149 |
+
"loss": 241.9411376953125,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.02093613301772113,
|
| 154 |
+
"grad_norm": 63.37522888183594,
|
| 155 |
+
"learning_rate": 4.999139182552709e-05,
|
| 156 |
+
"loss": 249.7280517578125,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.021982939668607183,
|
| 161 |
+
"grad_norm": 60.007144927978516,
|
| 162 |
+
"learning_rate": 4.9990507278185374e-05,
|
| 163 |
+
"loss": 236.941015625,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.02302974631949324,
|
| 168 |
+
"grad_norm": 66.06304168701172,
|
| 169 |
+
"learning_rate": 4.998957948790677e-05,
|
| 170 |
+
"loss": 240.55878363715277,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.0240765529703793,
|
| 175 |
+
"grad_norm": 68.9736099243164,
|
| 176 |
+
"learning_rate": 4.998860845629671e-05,
|
| 177 |
+
"loss": 247.988720703125,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.025123359621265353,
|
| 182 |
+
"grad_norm": 61.50578308105469,
|
| 183 |
+
"learning_rate": 4.998759418503544e-05,
|
| 184 |
+
"loss": 233.025537109375,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.02617016627215141,
|
| 189 |
+
"grad_norm": 54.96660614013672,
|
| 190 |
+
"learning_rate": 4.998653667587801e-05,
|
| 191 |
+
"loss": 243.004736328125,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.06804234327362181,
|
| 196 |
+
"grad_norm": 50.733734130859375,
|
| 197 |
+
"learning_rate": 4.990903999473118e-05,
|
| 198 |
+
"loss": 250.31749378551137,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.07065935647645342,
|
| 203 |
+
"grad_norm": 48.06721496582031,
|
| 204 |
+
"learning_rate": 4.9901899191179905e-05,
|
| 205 |
+
"loss": 243.69541015625,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.07327636967928504,
|
| 210 |
+
"grad_norm": 54.863311767578125,
|
| 211 |
+
"learning_rate": 4.989448913413268e-05,
|
| 212 |
+
"loss": 225.0319580078125,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07589338288211664,
|
| 217 |
+
"grad_norm": 60.36458969116211,
|
| 218 |
+
"learning_rate": 4.988680990371125e-05,
|
| 219 |
+
"loss": 222.7537353515625,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.07851039608494825,
|
| 224 |
+
"grad_norm": 57.877986907958984,
|
| 225 |
+
"learning_rate": 4.987886158294783e-05,
|
| 226 |
+
"loss": 241.103515625,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.08112740928777985,
|
| 231 |
+
"grad_norm": 64.6783676147461,
|
| 232 |
+
"learning_rate": 4.987064425778417e-05,
|
| 233 |
+
"loss": 215.971142578125,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.08374442249061147,
|
| 238 |
+
"grad_norm": 50.828765869140625,
|
| 239 |
+
"learning_rate": 4.986215801707066e-05,
|
| 240 |
+
"loss": 213.7265380859375,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.08636143569344307,
|
| 245 |
+
"grad_norm": 55.68048858642578,
|
| 246 |
+
"learning_rate": 4.9853402952565383e-05,
|
| 247 |
+
"loss": 216.94306640625,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.08897844889627468,
|
| 252 |
+
"grad_norm": 54.125267028808594,
|
| 253 |
+
"learning_rate": 4.984437915893305e-05,
|
| 254 |
+
"loss": 204.139990234375,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.0915954620991063,
|
| 259 |
+
"grad_norm": 57.27487564086914,
|
| 260 |
+
"learning_rate": 4.9835086733744074e-05,
|
| 261 |
+
"loss": 202.9946533203125,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.0942124753019379,
|
| 266 |
+
"grad_norm": 65.67598724365234,
|
| 267 |
+
"learning_rate": 4.9825525777473424e-05,
|
| 268 |
+
"loss": 207.21923828125,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.09682948850476951,
|
| 273 |
+
"grad_norm": 71.1091537475586,
|
| 274 |
+
"learning_rate": 4.98156963934996e-05,
|
| 275 |
+
"loss": 204.4235107421875,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.09944650170760111,
|
| 280 |
+
"grad_norm": 60.16761779785156,
|
| 281 |
+
"learning_rate": 4.9805598688103496e-05,
|
| 282 |
+
"loss": 213.841748046875,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.19596246146437973,
|
| 287 |
+
"grad_norm": 268.7782897949219,
|
| 288 |
+
"learning_rate": 4.99999986487106e-05,
|
| 289 |
+
"loss": 474.80499267578125,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2009871399634664,
|
| 294 |
+
"grad_norm": 107.96406555175781,
|
| 295 |
+
"learning_rate": 4.9999459486180746e-05,
|
| 296 |
+
"loss": 528.41904296875,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.20601181846255306,
|
| 301 |
+
"grad_norm": 133.09701538085938,
|
| 302 |
+
"learning_rate": 4.9997837968095394e-05,
|
| 303 |
+
"loss": 484.614111328125,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.21103649696163973,
|
| 308 |
+
"grad_norm": 102.78153228759766,
|
| 309 |
+
"learning_rate": 4.999513551586019e-05,
|
| 310 |
+
"loss": 467.132373046875,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.21606117546072637,
|
| 315 |
+
"grad_norm": 92.18177795410156,
|
| 316 |
+
"learning_rate": 4.999135224633215e-05,
|
| 317 |
+
"loss": 463.333203125,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.22108585395981303,
|
| 322 |
+
"grad_norm": 106.00531005859375,
|
| 323 |
+
"learning_rate": 4.998648832310402e-05,
|
| 324 |
+
"loss": 452.96220703125,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2261105324588997,
|
| 329 |
+
"grad_norm": 100.12345123291016,
|
| 330 |
+
"learning_rate": 4.9980543956497236e-05,
|
| 331 |
+
"loss": 425.08818359375,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.23113521095798636,
|
| 336 |
+
"grad_norm": 96.5484619140625,
|
| 337 |
+
"learning_rate": 4.997351940355277e-05,
|
| 338 |
+
"loss": 474.501953125,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.23615988945707303,
|
| 343 |
+
"grad_norm": 104.5451889038086,
|
| 344 |
+
"learning_rate": 4.996541496802006e-05,
|
| 345 |
+
"loss": 447.30810546875,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.24118456795615967,
|
| 350 |
+
"grad_norm": 129.78697204589844,
|
| 351 |
+
"learning_rate": 4.995623100034386e-05,
|
| 352 |
+
"loss": 442.8482421875,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.24620924645524633,
|
| 357 |
+
"grad_norm": 95.0133285522461,
|
| 358 |
+
"learning_rate": 4.994596789764909e-05,
|
| 359 |
+
"loss": 433.66689453125,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.251233924954333,
|
| 364 |
+
"grad_norm": 93.045654296875,
|
| 365 |
+
"learning_rate": 4.993462610372364e-05,
|
| 366 |
+
"loss": 418.920263671875,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.25625860345341966,
|
| 371 |
+
"grad_norm": 102.6265869140625,
|
| 372 |
+
"learning_rate": 4.992220610899924e-05,
|
| 373 |
+
"loss": 394.33173828125,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.26128328195250633,
|
| 378 |
+
"grad_norm": 94.07565307617188,
|
| 379 |
+
"learning_rate": 4.990870845053017e-05,
|
| 380 |
+
"loss": 388.1531494140625,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.266307960451593,
|
| 385 |
+
"grad_norm": 105.10103607177734,
|
| 386 |
+
"learning_rate": 4.989413371197013e-05,
|
| 387 |
+
"loss": 379.9688720703125,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.27133263895067966,
|
| 392 |
+
"grad_norm": 99.91519927978516,
|
| 393 |
+
"learning_rate": 4.987848252354691e-05,
|
| 394 |
+
"loss": 388.9978271484375,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.2763573174497663,
|
| 399 |
+
"grad_norm": 93.43067169189453,
|
| 400 |
+
"learning_rate": 4.986175556203521e-05,
|
| 401 |
+
"loss": 359.740380859375,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.28138199594885294,
|
| 406 |
+
"grad_norm": 86.09584045410156,
|
| 407 |
+
"learning_rate": 4.984395355072732e-05,
|
| 408 |
+
"loss": 394.5279296875,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.2864066744479396,
|
| 413 |
+
"grad_norm": 92.78561401367188,
|
| 414 |
+
"learning_rate": 4.982507725940191e-05,
|
| 415 |
+
"loss": 379.4552490234375,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.29143135294702627,
|
| 420 |
+
"grad_norm": 89.77680969238281,
|
| 421 |
+
"learning_rate": 4.980512750429068e-05,
|
| 422 |
+
"loss": 382.2197509765625,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.29645603144611293,
|
| 427 |
+
"grad_norm": 90.80536651611328,
|
| 428 |
+
"learning_rate": 4.97841051480431e-05,
|
| 429 |
+
"loss": 322.7532958984375,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3014807099451996,
|
| 434 |
+
"grad_norm": 95.0079574584961,
|
| 435 |
+
"learning_rate": 4.976201109968908e-05,
|
| 436 |
+
"loss": 383.434130859375,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.30650538844428626,
|
| 441 |
+
"grad_norm": 96.08088684082031,
|
| 442 |
+
"learning_rate": 4.973884631459971e-05,
|
| 443 |
+
"loss": 315.9826904296875,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.31153006694337293,
|
| 448 |
+
"grad_norm": 95.79961395263672,
|
| 449 |
+
"learning_rate": 4.9714611794445906e-05,
|
| 450 |
+
"loss": 300.992431640625,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3165547454424596,
|
| 455 |
+
"grad_norm": 85.15829467773438,
|
| 456 |
+
"learning_rate": 4.968930858715511e-05,
|
| 457 |
+
"loss": 345.0525390625,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.32157942394154626,
|
| 462 |
+
"grad_norm": 97.35791778564453,
|
| 463 |
+
"learning_rate": 4.966293778686598e-05,
|
| 464 |
+
"loss": 303.704541015625,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3266041024406329,
|
| 469 |
+
"grad_norm": 100.1771469116211,
|
| 470 |
+
"learning_rate": 4.963550053388106e-05,
|
| 471 |
+
"loss": 326.5158447265625,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.33162878093971954,
|
| 476 |
+
"grad_norm": 90.39754486083984,
|
| 477 |
+
"learning_rate": 4.9606998014617535e-05,
|
| 478 |
+
"loss": 307.4456787109375,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3366534594388062,
|
| 483 |
+
"grad_norm": 95.20532989501953,
|
| 484 |
+
"learning_rate": 4.957743146155581e-05,
|
| 485 |
+
"loss": 268.6821533203125,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.34167813793789287,
|
| 490 |
+
"grad_norm": 79.56786346435547,
|
| 491 |
+
"learning_rate": 4.954680215318635e-05,
|
| 492 |
+
"loss": 271.5860107421875,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.34670281643697953,
|
| 497 |
+
"grad_norm": 70.59774017333984,
|
| 498 |
+
"learning_rate": 4.951511141395432e-05,
|
| 499 |
+
"loss": 274.2177734375,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3517274949360662,
|
| 504 |
+
"grad_norm": 85.6360092163086,
|
| 505 |
+
"learning_rate": 4.948236061420229e-05,
|
| 506 |
+
"loss": 282.2251708984375,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.35675217343515286,
|
| 511 |
+
"grad_norm": 80.13080596923828,
|
| 512 |
+
"learning_rate": 4.944855117011107e-05,
|
| 513 |
+
"loss": 288.8053466796875,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.36177685193423953,
|
| 518 |
+
"grad_norm": 72.03929901123047,
|
| 519 |
+
"learning_rate": 4.941368454363839e-05,
|
| 520 |
+
"loss": 260.295849609375,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3668015304333262,
|
| 525 |
+
"grad_norm": 71.65677642822266,
|
| 526 |
+
"learning_rate": 4.937776224245573e-05,
|
| 527 |
+
"loss": 283.718701171875,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.37182620893241286,
|
| 532 |
+
"grad_norm": 79.88951110839844,
|
| 533 |
+
"learning_rate": 4.934078581988311e-05,
|
| 534 |
+
"loss": 237.0281494140625,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3768508874314995,
|
| 539 |
+
"grad_norm": 70.0428237915039,
|
| 540 |
+
"learning_rate": 4.930275687482191e-05,
|
| 541 |
+
"loss": 236.0194091796875,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.38187556593058614,
|
| 546 |
+
"grad_norm": 92.4908447265625,
|
| 547 |
+
"learning_rate": 4.926367705168577e-05,
|
| 548 |
+
"loss": 240.9611572265625,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.3869002444296728,
|
| 553 |
+
"grad_norm": 70.08905792236328,
|
| 554 |
+
"learning_rate": 4.922354804032944e-05,
|
| 555 |
+
"loss": 230.3265380859375,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.39192492292875947,
|
| 560 |
+
"grad_norm": 77.18536376953125,
|
| 561 |
+
"learning_rate": 4.9182371575975736e-05,
|
| 562 |
+
"loss": 242.0444580078125,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.39694960142784613,
|
| 567 |
+
"grad_norm": 79.4393081665039,
|
| 568 |
+
"learning_rate": 4.9140149439140506e-05,
|
| 569 |
+
"loss": 247.16416015625,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.4019742799269328,
|
| 574 |
+
"grad_norm": 70.19295501708984,
|
| 575 |
+
"learning_rate": 4.909688345555561e-05,
|
| 576 |
+
"loss": 221.3811279296875,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.40699895842601946,
|
| 581 |
+
"grad_norm": 66.46074676513672,
|
| 582 |
+
"learning_rate": 4.9052575496090016e-05,
|
| 583 |
+
"loss": 222.3531005859375,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.41202363692510613,
|
| 588 |
+
"grad_norm": 103.58501434326172,
|
| 589 |
+
"learning_rate": 4.900722747666888e-05,
|
| 590 |
+
"loss": 237.266455078125,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.4170483154241928,
|
| 595 |
+
"grad_norm": 84.74609375,
|
| 596 |
+
"learning_rate": 4.8960841358190704e-05,
|
| 597 |
+
"loss": 220.35146484375,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.42207299392327946,
|
| 602 |
+
"grad_norm": 75.79342651367188,
|
| 603 |
+
"learning_rate": 4.891341914644251e-05,
|
| 604 |
+
"loss": 207.2835693359375,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4270976724223661,
|
| 609 |
+
"grad_norm": 84.80883026123047,
|
| 610 |
+
"learning_rate": 4.886496289201318e-05,
|
| 611 |
+
"loss": 199.74576416015626,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.43212235092145274,
|
| 616 |
+
"grad_norm": 67.72750091552734,
|
| 617 |
+
"learning_rate": 4.881547469020471e-05,
|
| 618 |
+
"loss": 198.93392333984374,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4371470294205394,
|
| 623 |
+
"grad_norm": 76.4952163696289,
|
| 624 |
+
"learning_rate": 4.8764956680941676e-05,
|
| 625 |
+
"loss": 189.1037109375,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.44217170791962607,
|
| 630 |
+
"grad_norm": 78.63851928710938,
|
| 631 |
+
"learning_rate": 4.8713411048678635e-05,
|
| 632 |
+
"loss": 226.8509521484375,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.44719638641871273,
|
| 637 |
+
"grad_norm": 63.36662292480469,
|
| 638 |
+
"learning_rate": 4.866084002230573e-05,
|
| 639 |
+
"loss": 202.01885986328125,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4522210649177994,
|
| 644 |
+
"grad_norm": 82.46707916259766,
|
| 645 |
+
"learning_rate": 4.860724587505223e-05,
|
| 646 |
+
"loss": 190.03756103515624,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.45724574341688606,
|
| 651 |
+
"grad_norm": 79.09195709228516,
|
| 652 |
+
"learning_rate": 4.8552630924388335e-05,
|
| 653 |
+
"loss": 193.71884765625,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.46227042191597273,
|
| 658 |
+
"grad_norm": 72.94918060302734,
|
| 659 |
+
"learning_rate": 4.8496997531924885e-05,
|
| 660 |
+
"loss": 174.4781982421875,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4672951004150594,
|
| 665 |
+
"grad_norm": 97.54232025146484,
|
| 666 |
+
"learning_rate": 4.8440348103311276e-05,
|
| 667 |
+
"loss": 196.74876708984374,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.47231977891414606,
|
| 672 |
+
"grad_norm": 70.02765655517578,
|
| 673 |
+
"learning_rate": 4.838268508813143e-05,
|
| 674 |
+
"loss": 183.72100830078125,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.4773444574132327,
|
| 679 |
+
"grad_norm": 62.996543884277344,
|
| 680 |
+
"learning_rate": 4.8324010979797875e-05,
|
| 681 |
+
"loss": 171.5578369140625,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.48236913591231934,
|
| 686 |
+
"grad_norm": 72.1956787109375,
|
| 687 |
+
"learning_rate": 4.8264328315443914e-05,
|
| 688 |
+
"loss": 167.999951171875,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.487393814411406,
|
| 693 |
+
"grad_norm": 96.08892822265625,
|
| 694 |
+
"learning_rate": 4.820363967581394e-05,
|
| 695 |
+
"loss": 159.68211669921874,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.49241849291049267,
|
| 700 |
+
"grad_norm": 65.87628173828125,
|
| 701 |
+
"learning_rate": 4.814194768515182e-05,
|
| 702 |
+
"loss": 156.25810546875,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.49744317140957933,
|
| 707 |
+
"grad_norm": 70.53435516357422,
|
| 708 |
+
"learning_rate": 4.807925501108744e-05,
|
| 709 |
+
"loss": 158.4431640625,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.502467849908666,
|
| 714 |
+
"grad_norm": 71.5821304321289,
|
| 715 |
+
"learning_rate": 4.801556436452133e-05,
|
| 716 |
+
"loss": 157.15218505859374,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5074925284077526,
|
| 721 |
+
"grad_norm": 62.593109130859375,
|
| 722 |
+
"learning_rate": 4.795087849950747e-05,
|
| 723 |
+
"loss": 154.85384521484374,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5125172069068393,
|
| 728 |
+
"grad_norm": 70.64319610595703,
|
| 729 |
+
"learning_rate": 4.7885200213134164e-05,
|
| 730 |
+
"loss": 144.212548828125,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5175418854059259,
|
| 735 |
+
"grad_norm": 57.551979064941406,
|
| 736 |
+
"learning_rate": 4.781853234540312e-05,
|
| 737 |
+
"loss": 150.35511474609376,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5225665639050127,
|
| 742 |
+
"grad_norm": 60.21275329589844,
|
| 743 |
+
"learning_rate": 4.7750877779106666e-05,
|
| 744 |
+
"loss": 139.1343994140625,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5275912424040993,
|
| 749 |
+
"grad_norm": 74.57681274414062,
|
| 750 |
+
"learning_rate": 4.7682239439703025e-05,
|
| 751 |
+
"loss": 152.9317138671875,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.532615920903186,
|
| 756 |
+
"grad_norm": 62.081398010253906,
|
| 757 |
+
"learning_rate": 4.761262029518988e-05,
|
| 758 |
+
"loss": 138.762646484375,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5376405994022726,
|
| 763 |
+
"grad_norm": 97.3165283203125,
|
| 764 |
+
"learning_rate": 4.754202335597601e-05,
|
| 765 |
+
"loss": 132.7292236328125,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5426652779013593,
|
| 770 |
+
"grad_norm": 68.6803970336914,
|
| 771 |
+
"learning_rate": 4.7470451674751106e-05,
|
| 772 |
+
"loss": 129.36715087890624,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5476899564004459,
|
| 777 |
+
"grad_norm": 63.835575103759766,
|
| 778 |
+
"learning_rate": 4.7397908346353796e-05,
|
| 779 |
+
"loss": 127.2830078125,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.5527146348995327,
|
| 784 |
+
"grad_norm": 63.419559478759766,
|
| 785 |
+
"learning_rate": 4.73243965076378e-05,
|
| 786 |
+
"loss": 129.15245361328124,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5577393133986193,
|
| 791 |
+
"grad_norm": 53.640995025634766,
|
| 792 |
+
"learning_rate": 4.724991933733629e-05,
|
| 793 |
+
"loss": 122.0046630859375,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5627639918977059,
|
| 798 |
+
"grad_norm": 59.515647888183594,
|
| 799 |
+
"learning_rate": 4.7174480055924455e-05,
|
| 800 |
+
"loss": 135.0306396484375,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5677886703967926,
|
| 805 |
+
"grad_norm": 61.889373779296875,
|
| 806 |
+
"learning_rate": 4.7098081925480216e-05,
|
| 807 |
+
"loss": 123.96072998046876,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5728133488958792,
|
| 812 |
+
"grad_norm": 52.50545120239258,
|
| 813 |
+
"learning_rate": 4.7020728249543196e-05,
|
| 814 |
+
"loss": 120.4759521484375,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.5778380273949659,
|
| 819 |
+
"grad_norm": 64.7458267211914,
|
| 820 |
+
"learning_rate": 4.694242237297186e-05,
|
| 821 |
+
"loss": 131.4004150390625,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.5828627058940525,
|
| 826 |
+
"grad_norm": 52.680599212646484,
|
| 827 |
+
"learning_rate": 4.686316768179889e-05,
|
| 828 |
+
"loss": 124.689404296875,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.5878873843931393,
|
| 833 |
+
"grad_norm": 77.16497039794922,
|
| 834 |
+
"learning_rate": 4.678296760308474e-05,
|
| 835 |
+
"loss": 120.44024658203125,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.5929120628922259,
|
| 840 |
+
"grad_norm": 60.93242263793945,
|
| 841 |
+
"learning_rate": 4.670182560476948e-05,
|
| 842 |
+
"loss": 120.22978515625,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.5979367413913126,
|
| 847 |
+
"grad_norm": 59.64492416381836,
|
| 848 |
+
"learning_rate": 4.661974519552283e-05,
|
| 849 |
+
"loss": 108.0113525390625,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6029614198903992,
|
| 854 |
+
"grad_norm": 65.3152847290039,
|
| 855 |
+
"learning_rate": 4.653672992459242e-05,
|
| 856 |
+
"loss": 112.077978515625,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.6079860983894858,
|
| 861 |
+
"grad_norm": 58.9239616394043,
|
| 862 |
+
"learning_rate": 4.645278338165035e-05,
|
| 863 |
+
"loss": 106.50635986328125,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.6130107768885725,
|
| 868 |
+
"grad_norm": 56.77158737182617,
|
| 869 |
+
"learning_rate": 4.636790919663794e-05,
|
| 870 |
+
"loss": 102.1367431640625,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.6180354553876591,
|
| 875 |
+
"grad_norm": 49.77719497680664,
|
| 876 |
+
"learning_rate": 4.6282111039608784e-05,
|
| 877 |
+
"loss": 97.8525634765625,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.6230601338867459,
|
| 882 |
+
"grad_norm": 51.9918212890625,
|
| 883 |
+
"learning_rate": 4.6195392620570035e-05,
|
| 884 |
+
"loss": 105.98614501953125,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.6280848123858325,
|
| 889 |
+
"grad_norm": 58.79019546508789,
|
| 890 |
+
"learning_rate": 4.6107757689322014e-05,
|
| 891 |
+
"loss": 102.065966796875,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.6331094908849192,
|
| 896 |
+
"grad_norm": 49.530391693115234,
|
| 897 |
+
"learning_rate": 4.6019210035296034e-05,
|
| 898 |
+
"loss": 98.87720947265625,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.6381341693840058,
|
| 903 |
+
"grad_norm": 53.472782135009766,
|
| 904 |
+
"learning_rate": 4.592975348739054e-05,
|
| 905 |
+
"loss": 108.6241943359375,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.6431588478830925,
|
| 910 |
+
"grad_norm": 63.10646438598633,
|
| 911 |
+
"learning_rate": 4.584393142668341e-05,
|
| 912 |
+
"loss": 93.73190307617188,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.6481835263821791,
|
| 917 |
+
"grad_norm": 59.84651565551758,
|
| 918 |
+
"learning_rate": 4.5752713697269135e-05,
|
| 919 |
+
"loss": 96.8425048828125,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.6532082048812659,
|
| 924 |
+
"grad_norm": 60.635169982910156,
|
| 925 |
+
"learning_rate": 4.5660598597571637e-05,
|
| 926 |
+
"loss": 103.2667236328125,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.6582328833803525,
|
| 931 |
+
"grad_norm": 61.401737213134766,
|
| 932 |
+
"learning_rate": 4.5567590110749646e-05,
|
| 933 |
+
"loss": 87.72919311523438,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.6632575618794391,
|
| 938 |
+
"grad_norm": 60.122772216796875,
|
| 939 |
+
"learning_rate": 4.547369225859297e-05,
|
| 940 |
+
"loss": 94.103662109375,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.6682822403785258,
|
| 945 |
+
"grad_norm": 55.90855026245117,
|
| 946 |
+
"learning_rate": 4.537890910134854e-05,
|
| 947 |
+
"loss": 83.04338989257812,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.6733069188776124,
|
| 952 |
+
"grad_norm": 71.83133697509766,
|
| 953 |
+
"learning_rate": 4.528324473754486e-05,
|
| 954 |
+
"loss": 83.29258422851562,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.6783315973766991,
|
| 959 |
+
"grad_norm": 58.1335334777832,
|
| 960 |
+
"learning_rate": 4.518670330381479e-05,
|
| 961 |
+
"loss": 81.46260986328124,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.6833562758757857,
|
| 966 |
+
"grad_norm": 73.3138656616211,
|
| 967 |
+
"learning_rate": 4.5089288974716634e-05,
|
| 968 |
+
"loss": 81.26986694335938,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.6883809543748725,
|
| 973 |
+
"grad_norm": 64.87886810302734,
|
| 974 |
+
"learning_rate": 4.49910059625537e-05,
|
| 975 |
+
"loss": 82.78674926757813,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.6934056328739591,
|
| 980 |
+
"grad_norm": 57.293861389160156,
|
| 981 |
+
"learning_rate": 4.489185851719208e-05,
|
| 982 |
+
"loss": 82.71456909179688,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.6984303113730458,
|
| 987 |
+
"grad_norm": 53.40017318725586,
|
| 988 |
+
"learning_rate": 4.4791850925876926e-05,
|
| 989 |
+
"loss": 84.79779052734375,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.7034549898721324,
|
| 994 |
+
"grad_norm": 48.285736083984375,
|
| 995 |
+
"learning_rate": 4.4690987513047044e-05,
|
| 996 |
+
"loss": 81.11317138671875,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.708479668371219,
|
| 1001 |
+
"grad_norm": 56.2747802734375,
|
| 1002 |
+
"learning_rate": 4.4589272640147915e-05,
|
| 1003 |
+
"loss": 72.64208374023437,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.7135043468703057,
|
| 1008 |
+
"grad_norm": 54.770931243896484,
|
| 1009 |
+
"learning_rate": 4.44867107054431e-05,
|
| 1010 |
+
"loss": 75.95457153320312,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.7185290253693923,
|
| 1015 |
+
"grad_norm": 49.85646057128906,
|
| 1016 |
+
"learning_rate": 4.4383306143824035e-05,
|
| 1017 |
+
"loss": 74.01212158203126,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.7235537038684791,
|
| 1022 |
+
"grad_norm": 49.286746978759766,
|
| 1023 |
+
"learning_rate": 4.427906342661828e-05,
|
| 1024 |
+
"loss": 67.97941284179687,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.7285783823675657,
|
| 1029 |
+
"grad_norm": 56.60735321044922,
|
| 1030 |
+
"learning_rate": 4.4173987061396186e-05,
|
| 1031 |
+
"loss": 70.84263916015625,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.7336030608666524,
|
| 1036 |
+
"grad_norm": 57.23542404174805,
|
| 1037 |
+
"learning_rate": 4.406808159177593e-05,
|
| 1038 |
+
"loss": 65.6290283203125,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.738627739365739,
|
| 1043 |
+
"grad_norm": 83.55549621582031,
|
| 1044 |
+
"learning_rate": 4.396135159722711e-05,
|
| 1045 |
+
"loss": 65.52483520507812,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.7436524178648257,
|
| 1050 |
+
"grad_norm": 53.960811614990234,
|
| 1051 |
+
"learning_rate": 4.3853801692872686e-05,
|
| 1052 |
+
"loss": 67.09555053710938,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.7486770963639123,
|
| 1057 |
+
"grad_norm": 45.474632263183594,
|
| 1058 |
+
"learning_rate": 4.374543652928943e-05,
|
| 1059 |
+
"loss": 67.3612548828125,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.753701774862999,
|
| 1064 |
+
"grad_norm": 56.11079025268555,
|
| 1065 |
+
"learning_rate": 4.363626079230681e-05,
|
| 1066 |
+
"loss": 62.279473876953126,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.7587264533620857,
|
| 1071 |
+
"grad_norm": 49.5059928894043,
|
| 1072 |
+
"learning_rate": 4.3526279202804387e-05,
|
| 1073 |
+
"loss": 73.84429321289062,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.7637511318611723,
|
| 1078 |
+
"grad_norm": 53.92646789550781,
|
| 1079 |
+
"learning_rate": 4.3415496516507695e-05,
|
| 1080 |
+
"loss": 61.59212646484375,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.768775810360259,
|
| 1085 |
+
"grad_norm": 57.92325210571289,
|
| 1086 |
+
"learning_rate": 4.3303917523782546e-05,
|
| 1087 |
+
"loss": 65.36521606445312,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.7738004888593456,
|
| 1092 |
+
"grad_norm": 47.250648498535156,
|
| 1093 |
+
"learning_rate": 4.3191547049427954e-05,
|
| 1094 |
+
"loss": 66.87857055664062,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.7788251673584323,
|
| 1099 |
+
"grad_norm": 91.52741241455078,
|
| 1100 |
+
"learning_rate": 4.307838995246745e-05,
|
| 1101 |
+
"loss": 62.80021362304687,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.7838498458575189,
|
| 1106 |
+
"grad_norm": 46.733856201171875,
|
| 1107 |
+
"learning_rate": 4.296445112593902e-05,
|
| 1108 |
+
"loss": 55.04381103515625,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.7888745243566057,
|
| 1113 |
+
"grad_norm": 44.601497650146484,
|
| 1114 |
+
"learning_rate": 4.284973549668347e-05,
|
| 1115 |
+
"loss": 55.08817749023437,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.7938992028556923,
|
| 1120 |
+
"grad_norm": 50.94633102416992,
|
| 1121 |
+
"learning_rate": 4.273424802513145e-05,
|
| 1122 |
+
"loss": 52.93193359375,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.798923881354779,
|
| 1127 |
+
"grad_norm": 43.45872497558594,
|
| 1128 |
+
"learning_rate": 4.2617993705088886e-05,
|
| 1129 |
+
"loss": 54.57942504882813,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.8039485598538656,
|
| 1134 |
+
"grad_norm": 46.4177131652832,
|
| 1135 |
+
"learning_rate": 4.2500977563521115e-05,
|
| 1136 |
+
"loss": 54.19765014648438,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.8089732383529522,
|
| 1141 |
+
"grad_norm": 48.374786376953125,
|
| 1142 |
+
"learning_rate": 4.238320466033546e-05,
|
| 1143 |
+
"loss": 51.5730712890625,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.8139979168520389,
|
| 1148 |
+
"grad_norm": 51.431087493896484,
|
| 1149 |
+
"learning_rate": 4.2264680088162456e-05,
|
| 1150 |
+
"loss": 52.94944458007812,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.8190225953511255,
|
| 1155 |
+
"grad_norm": 40.70262145996094,
|
| 1156 |
+
"learning_rate": 4.214540897213565e-05,
|
| 1157 |
+
"loss": 53.7551513671875,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.8240472738502123,
|
| 1162 |
+
"grad_norm": 43.26173782348633,
|
| 1163 |
+
"learning_rate": 4.202539646966993e-05,
|
| 1164 |
+
"loss": 55.24525756835938,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.8290719523492989,
|
| 1169 |
+
"grad_norm": 47.32640838623047,
|
| 1170 |
+
"learning_rate": 4.190464777023861e-05,
|
| 1171 |
+
"loss": 48.86954040527344,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.8340966308483856,
|
| 1176 |
+
"grad_norm": 47.30034255981445,
|
| 1177 |
+
"learning_rate": 4.178316809514893e-05,
|
| 1178 |
+
"loss": 46.55088806152344,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.8391213093474722,
|
| 1183 |
+
"grad_norm": 51.20639419555664,
|
| 1184 |
+
"learning_rate": 4.1660962697316346e-05,
|
| 1185 |
+
"loss": 48.73444519042969,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.8441459878465589,
|
| 1190 |
+
"grad_norm": 51.74330139160156,
|
| 1191 |
+
"learning_rate": 4.153803686103737e-05,
|
| 1192 |
+
"loss": 45.40309448242188,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.8491706663456455,
|
| 1197 |
+
"grad_norm": 41.944496154785156,
|
| 1198 |
+
"learning_rate": 4.141439590176105e-05,
|
| 1199 |
+
"loss": 46.365499877929686,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.8541953448447323,
|
| 1204 |
+
"grad_norm": 41.049713134765625,
|
| 1205 |
+
"learning_rate": 4.129004516585916e-05,
|
| 1206 |
+
"loss": 47.59626770019531,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.8592200233438189,
|
| 1211 |
+
"grad_norm": 44.36544418334961,
|
| 1212 |
+
"learning_rate": 4.116499003039499e-05,
|
| 1213 |
+
"loss": 42.41243286132813,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.8642447018429055,
|
| 1218 |
+
"grad_norm": 46.1484489440918,
|
| 1219 |
+
"learning_rate": 4.1039235902890875e-05,
|
| 1220 |
+
"loss": 44.016815185546875,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.8692693803419922,
|
| 1225 |
+
"grad_norm": 38.971961975097656,
|
| 1226 |
+
"learning_rate": 4.091278822109428e-05,
|
| 1227 |
+
"loss": 41.796511840820315,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.8742940588410788,
|
| 1232 |
+
"grad_norm": 54.43899917602539,
|
| 1233 |
+
"learning_rate": 4.0785652452742776e-05,
|
| 1234 |
+
"loss": 43.20156555175781,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.8793187373401655,
|
| 1239 |
+
"grad_norm": 36.39228820800781,
|
| 1240 |
+
"learning_rate": 4.065783409532755e-05,
|
| 1241 |
+
"loss": 41.3662353515625,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.8843434158392521,
|
| 1246 |
+
"grad_norm": 43.41612243652344,
|
| 1247 |
+
"learning_rate": 4.052933867585567e-05,
|
| 1248 |
+
"loss": 41.19517211914062,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.8893680943383389,
|
| 1253 |
+
"grad_norm": 50.1053352355957,
|
| 1254 |
+
"learning_rate": 4.040017175061113e-05,
|
| 1255 |
+
"loss": 45.48277893066406,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.8943927728374255,
|
| 1260 |
+
"grad_norm": 68.26251983642578,
|
| 1261 |
+
"learning_rate": 4.027033890491458e-05,
|
| 1262 |
+
"loss": 40.87652587890625,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.8994174513365122,
|
| 1267 |
+
"grad_norm": 46.13456344604492,
|
| 1268 |
+
"learning_rate": 4.0139845752881814e-05,
|
| 1269 |
+
"loss": 39.160336303710935,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.9044421298355988,
|
| 1274 |
+
"grad_norm": 44.120697021484375,
|
| 1275 |
+
"learning_rate": 4.0008697937180984e-05,
|
| 1276 |
+
"loss": 41.72444458007813,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.9094668083346854,
|
| 1281 |
+
"grad_norm": 36.38677215576172,
|
| 1282 |
+
"learning_rate": 3.987690112878862e-05,
|
| 1283 |
+
"loss": 37.204827880859376,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.9144914868337721,
|
| 1288 |
+
"grad_norm": 49.02151870727539,
|
| 1289 |
+
"learning_rate": 3.974446102674444e-05,
|
| 1290 |
+
"loss": 37.10683898925781,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.9195161653328587,
|
| 1295 |
+
"grad_norm": 36.17109298706055,
|
| 1296 |
+
"learning_rate": 3.9611383357904866e-05,
|
| 1297 |
+
"loss": 38.15001525878906,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.9245408438319455,
|
| 1302 |
+
"grad_norm": 40.26628112792969,
|
| 1303 |
+
"learning_rate": 3.947767387669542e-05,
|
| 1304 |
+
"loss": 44.486627197265626,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.9295655223310321,
|
| 1309 |
+
"grad_norm": 38.918766021728516,
|
| 1310 |
+
"learning_rate": 3.934333836486189e-05,
|
| 1311 |
+
"loss": 35.51749267578125,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.9345902008301188,
|
| 1316 |
+
"grad_norm": 36.1772346496582,
|
| 1317 |
+
"learning_rate": 3.920838263122033e-05,
|
| 1318 |
+
"loss": 32.570498657226565,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.9396148793292054,
|
| 1323 |
+
"grad_norm": 39.06047058105469,
|
| 1324 |
+
"learning_rate": 3.9072812511405855e-05,
|
| 1325 |
+
"loss": 34.915673828125,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.9446395578282921,
|
| 1330 |
+
"grad_norm": 41.863502502441406,
|
| 1331 |
+
"learning_rate": 3.893663386762032e-05,
|
| 1332 |
+
"loss": 34.646026611328125,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.9496642363273787,
|
| 1337 |
+
"grad_norm": 32.75388717651367,
|
| 1338 |
+
"learning_rate": 3.8799852588378836e-05,
|
| 1339 |
+
"loss": 33.01146545410156,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.9546889148264655,
|
| 1344 |
+
"grad_norm": 45.868865966796875,
|
| 1345 |
+
"learning_rate": 3.8662474588255146e-05,
|
| 1346 |
+
"loss": 30.23636474609375,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.9597135933255521,
|
| 1351 |
+
"grad_norm": 37.765052795410156,
|
| 1352 |
+
"learning_rate": 3.852450580762584e-05,
|
| 1353 |
+
"loss": 31.944390869140626,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.9647382718246387,
|
| 1358 |
+
"grad_norm": 33.98917007446289,
|
| 1359 |
+
"learning_rate": 3.8385952212413525e-05,
|
| 1360 |
+
"loss": 31.06385498046875,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.9697629503237254,
|
| 1365 |
+
"grad_norm": 32.69452667236328,
|
| 1366 |
+
"learning_rate": 3.824681979382883e-05,
|
| 1367 |
+
"loss": 30.422335815429687,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.974787628822812,
|
| 1372 |
+
"grad_norm": 47.07701110839844,
|
| 1373 |
+
"learning_rate": 3.810711456811136e-05,
|
| 1374 |
+
"loss": 30.911959838867187,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.9798123073218987,
|
| 1379 |
+
"grad_norm": 37.96084976196289,
|
| 1380 |
+
"learning_rate": 3.796684257626952e-05,
|
| 1381 |
+
"loss": 29.244085693359374,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.9848369858209853,
|
| 1386 |
+
"grad_norm": 32.45835494995117,
|
| 1387 |
+
"learning_rate": 3.7826009883819315e-05,
|
| 1388 |
+
"loss": 27.361874389648438,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.989861664320072,
|
| 1393 |
+
"grad_norm": 35.04416275024414,
|
| 1394 |
+
"learning_rate": 3.7684622580522055e-05,
|
| 1395 |
+
"loss": 27.51587829589844,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.9948863428191587,
|
| 1400 |
+
"grad_norm": 45.533992767333984,
|
| 1401 |
+
"learning_rate": 3.754268678012106e-05,
|
| 1402 |
+
"loss": 27.282977294921874,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.9999110213182454,
|
| 1407 |
+
"grad_norm": 37.77592086791992,
|
| 1408 |
+
"learning_rate": 3.740020862007724e-05,
|
| 1409 |
+
"loss": 27.77158203125,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.0047734445741323,
|
| 1414 |
+
"grad_norm": 29.9664306640625,
|
| 1415 |
+
"learning_rate": 3.7257194261303765e-05,
|
| 1416 |
+
"loss": 24.126179504394532,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.009798123073219,
|
| 1421 |
+
"grad_norm": 36.00257873535156,
|
| 1422 |
+
"learning_rate": 3.71136498878996e-05,
|
| 1423 |
+
"loss": 25.783734130859376,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.0148228015723058,
|
| 1428 |
+
"grad_norm": 71.59026336669922,
|
| 1429 |
+
"learning_rate": 3.696958170688215e-05,
|
| 1430 |
+
"loss": 24.485565185546875,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.0198474800713924,
|
| 1435 |
+
"grad_norm": 31.764368057250977,
|
| 1436 |
+
"learning_rate": 3.682499594791885e-05,
|
| 1437 |
+
"loss": 23.261607360839843,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.024872158570479,
|
| 1442 |
+
"grad_norm": 31.287395477294922,
|
| 1443 |
+
"learning_rate": 3.667989886305774e-05,
|
| 1444 |
+
"loss": 23.564991760253907,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.0298968370695656,
|
| 1449 |
+
"grad_norm": 29.753856658935547,
|
| 1450 |
+
"learning_rate": 3.653429672645721e-05,
|
| 1451 |
+
"loss": 25.142503356933595,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.0349215155686522,
|
| 1456 |
+
"grad_norm": 29.53838539123535,
|
| 1457 |
+
"learning_rate": 3.638819583411458e-05,
|
| 1458 |
+
"loss": 23.48755340576172,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.039946194067739,
|
| 1463 |
+
"grad_norm": 29.025772094726562,
|
| 1464 |
+
"learning_rate": 3.624160250359398e-05,
|
| 1465 |
+
"loss": 21.697846984863283,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.0449708725668256,
|
| 1470 |
+
"grad_norm": 31.875118255615234,
|
| 1471 |
+
"learning_rate": 3.609452307375308e-05,
|
| 1472 |
+
"loss": 24.43586883544922,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.0499955510659122,
|
| 1477 |
+
"grad_norm": 34.18278121948242,
|
| 1478 |
+
"learning_rate": 3.5946963904469024e-05,
|
| 1479 |
+
"loss": 21.297456359863283,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.0550202295649989,
|
| 1484 |
+
"grad_norm": 31.082683563232422,
|
| 1485 |
+
"learning_rate": 3.579893137636343e-05,
|
| 1486 |
+
"loss": 22.34284362792969,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.0600449080640857,
|
| 1491 |
+
"grad_norm": 31.194740295410156,
|
| 1492 |
+
"learning_rate": 3.565043189052647e-05,
|
| 1493 |
+
"loss": 23.093136596679688,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.0650695865631723,
|
| 1498 |
+
"grad_norm": 32.549827575683594,
|
| 1499 |
+
"learning_rate": 3.550147186824008e-05,
|
| 1500 |
+
"loss": 21.879423522949217,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.070094265062259,
|
| 1505 |
+
"grad_norm": 32.4059944152832,
|
| 1506 |
+
"learning_rate": 3.53520577507003e-05,
|
| 1507 |
+
"loss": 20.31920623779297,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.0751189435613455,
|
| 1512 |
+
"grad_norm": 29.14031219482422,
|
| 1513 |
+
"learning_rate": 3.520219599873877e-05,
|
| 1514 |
+
"loss": 21.056622314453126,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.0801436220604321,
|
| 1519 |
+
"grad_norm": 23.570371627807617,
|
| 1520 |
+
"learning_rate": 3.5051893092543306e-05,
|
| 1521 |
+
"loss": 19.163606262207033,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.085168300559519,
|
| 1526 |
+
"grad_norm": 30.20454216003418,
|
| 1527 |
+
"learning_rate": 3.490115553137776e-05,
|
| 1528 |
+
"loss": 21.274189758300782,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.0901929790586056,
|
| 1533 |
+
"grad_norm": 27.927400588989258,
|
| 1534 |
+
"learning_rate": 3.47499898333009e-05,
|
| 1535 |
+
"loss": 18.568623352050782,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.0952176575576922,
|
| 1540 |
+
"grad_norm": 31.376602172851562,
|
| 1541 |
+
"learning_rate": 3.459840253488465e-05,
|
| 1542 |
+
"loss": 19.727203369140625,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.1002423360567788,
|
| 1547 |
+
"grad_norm": 36.069026947021484,
|
| 1548 |
+
"learning_rate": 3.444640019093136e-05,
|
| 1549 |
+
"loss": 18.880654907226564,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.1052670145558656,
|
| 1554 |
+
"grad_norm": 33.837711334228516,
|
| 1555 |
+
"learning_rate": 3.429398937419044e-05,
|
| 1556 |
+
"loss": 18.870826721191406,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.1102916930549522,
|
| 1561 |
+
"grad_norm": 29.549623489379883,
|
| 1562 |
+
"learning_rate": 3.4141176675074104e-05,
|
| 1563 |
+
"loss": 18.073291015625,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.1153163715540388,
|
| 1568 |
+
"grad_norm": 24.568265914916992,
|
| 1569 |
+
"learning_rate": 3.39879687013724e-05,
|
| 1570 |
+
"loss": 19.973286437988282,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.1203410500531255,
|
| 1575 |
+
"grad_norm": 26.699716567993164,
|
| 1576 |
+
"learning_rate": 3.383437207796748e-05,
|
| 1577 |
+
"loss": 17.782139587402344,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.1253657285522123,
|
| 1582 |
+
"grad_norm": 28.5200252532959,
|
| 1583 |
+
"learning_rate": 3.3680393446547155e-05,
|
| 1584 |
+
"loss": 17.14067687988281,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.130390407051299,
|
| 1589 |
+
"grad_norm": 38.18730163574219,
|
| 1590 |
+
"learning_rate": 3.3526039465317675e-05,
|
| 1591 |
+
"loss": 18.262770080566405,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.1354150855503855,
|
| 1596 |
+
"grad_norm": 22.557958602905273,
|
| 1597 |
+
"learning_rate": 3.337131680871584e-05,
|
| 1598 |
+
"loss": 17.867935180664062,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.1404397640494721,
|
| 1603 |
+
"grad_norm": 22.901031494140625,
|
| 1604 |
+
"learning_rate": 3.3216232167120356e-05,
|
| 1605 |
+
"loss": 17.154512023925783,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.1454644425485587,
|
| 1610 |
+
"grad_norm": 27.419776916503906,
|
| 1611 |
+
"learning_rate": 3.30607922465626e-05,
|
| 1612 |
+
"loss": 16.06898193359375,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.1504891210476456,
|
| 1617 |
+
"grad_norm": 25.54627799987793,
|
| 1618 |
+
"learning_rate": 3.2905003768436576e-05,
|
| 1619 |
+
"loss": 15.820521545410156,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.1555137995467322,
|
| 1624 |
+
"grad_norm": 23.961387634277344,
|
| 1625 |
+
"learning_rate": 3.274887346920829e-05,
|
| 1626 |
+
"loss": 14.808900451660156,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.1605384780458188,
|
| 1631 |
+
"grad_norm": 23.095775604248047,
|
| 1632 |
+
"learning_rate": 3.259240810012452e-05,
|
| 1633 |
+
"loss": 14.714529418945313,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.1655631565449054,
|
| 1638 |
+
"grad_norm": 22.404417037963867,
|
| 1639 |
+
"learning_rate": 3.243561442692078e-05,
|
| 1640 |
+
"loss": 14.217619323730469,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.170587835043992,
|
| 1645 |
+
"grad_norm": 28.005577087402344,
|
| 1646 |
+
"learning_rate": 3.2278499229528836e-05,
|
| 1647 |
+
"loss": 15.66644287109375,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.1756125135430788,
|
| 1652 |
+
"grad_norm": 22.640522003173828,
|
| 1653 |
+
"learning_rate": 3.212106930178354e-05,
|
| 1654 |
+
"loss": 14.664166259765626,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.1806371920421654,
|
| 1659 |
+
"grad_norm": 24.132041931152344,
|
| 1660 |
+
"learning_rate": 3.1963331451129e-05,
|
| 1661 |
+
"loss": 16.09429473876953,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.185661870541252,
|
| 1666 |
+
"grad_norm": 25.478002548217773,
|
| 1667 |
+
"learning_rate": 3.180529249832428e-05,
|
| 1668 |
+
"loss": 14.240249633789062,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.1906865490403387,
|
| 1673 |
+
"grad_norm": 24.66731834411621,
|
| 1674 |
+
"learning_rate": 3.1646959277148406e-05,
|
| 1675 |
+
"loss": 14.484941101074218,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.1957112275394255,
|
| 1680 |
+
"grad_norm": 23.402143478393555,
|
| 1681 |
+
"learning_rate": 3.14883386341049e-05,
|
| 1682 |
+
"loss": 13.931251525878906,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.200735906038512,
|
| 1687 |
+
"grad_norm": 21.010387420654297,
|
| 1688 |
+
"learning_rate": 3.132943742812574e-05,
|
| 1689 |
+
"loss": 14.163758850097656,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.2057605845375987,
|
| 1694 |
+
"grad_norm": 20.641809463500977,
|
| 1695 |
+
"learning_rate": 3.1170262530274736e-05,
|
| 1696 |
+
"loss": 12.913584899902343,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.2107852630366853,
|
| 1701 |
+
"grad_norm": 23.429895401000977,
|
| 1702 |
+
"learning_rate": 3.101082082345045e-05,
|
| 1703 |
+
"loss": 12.281029510498048,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.2158099415357722,
|
| 1708 |
+
"grad_norm": 28.63577651977539,
|
| 1709 |
+
"learning_rate": 3.0851119202088555e-05,
|
| 1710 |
+
"loss": 12.448915100097656,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.2208346200348588,
|
| 1715 |
+
"grad_norm": 23.663249969482422,
|
| 1716 |
+
"learning_rate": 3.069116457186371e-05,
|
| 1717 |
+
"loss": 13.750531005859376,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.2258592985339454,
|
| 1722 |
+
"grad_norm": 22.31790542602539,
|
| 1723 |
+
"learning_rate": 3.053096384939097e-05,
|
| 1724 |
+
"loss": 12.410870361328126,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.230883977033032,
|
| 1729 |
+
"grad_norm": 21.297197341918945,
|
| 1730 |
+
"learning_rate": 3.037052396192667e-05,
|
| 1731 |
+
"loss": 12.693349456787109,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.2359086555321186,
|
| 1736 |
+
"grad_norm": 23.489055633544922,
|
| 1737 |
+
"learning_rate": 3.0209851847068932e-05,
|
| 1738 |
+
"loss": 12.602223968505859,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.2409333340312054,
|
| 1743 |
+
"grad_norm": 23.18564224243164,
|
| 1744 |
+
"learning_rate": 3.004895445245762e-05,
|
| 1745 |
+
"loss": 11.585443878173828,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.245958012530292,
|
| 1750 |
+
"grad_norm": 22.05788230895996,
|
| 1751 |
+
"learning_rate": 2.988783873547397e-05,
|
| 1752 |
+
"loss": 11.15335922241211,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.2509826910293786,
|
| 1757 |
+
"grad_norm": 21.006710052490234,
|
| 1758 |
+
"learning_rate": 2.9726511662939695e-05,
|
| 1759 |
+
"loss": 11.558931732177735,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.2560073695284653,
|
| 1764 |
+
"grad_norm": 33.19025421142578,
|
| 1765 |
+
"learning_rate": 2.956498021081575e-05,
|
| 1766 |
+
"loss": 11.444783020019532,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.2610320480275519,
|
| 1771 |
+
"grad_norm": 21.366567611694336,
|
| 1772 |
+
"learning_rate": 2.9403251363900736e-05,
|
| 1773 |
+
"loss": 10.560770416259766,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.2660567265266387,
|
| 1778 |
+
"grad_norm": 22.778701782226562,
|
| 1779 |
+
"learning_rate": 2.9241332115528763e-05,
|
| 1780 |
+
"loss": 10.666181945800782,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.2710814050257253,
|
| 1785 |
+
"grad_norm": 20.653833389282227,
|
| 1786 |
+
"learning_rate": 2.907922946726716e-05,
|
| 1787 |
+
"loss": 11.136424255371093,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.276106083524812,
|
| 1792 |
+
"grad_norm": 21.236427307128906,
|
| 1793 |
+
"learning_rate": 2.891695042861363e-05,
|
| 1794 |
+
"loss": 11.04461212158203,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.2811307620238988,
|
| 1799 |
+
"grad_norm": 39.907684326171875,
|
| 1800 |
+
"learning_rate": 2.875450201669323e-05,
|
| 1801 |
+
"loss": 10.980261993408202,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.2861554405229854,
|
| 1806 |
+
"grad_norm": 19.791812896728516,
|
| 1807 |
+
"learning_rate": 2.8591891255954873e-05,
|
| 1808 |
+
"loss": 10.691975402832032,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.291180119022072,
|
| 1813 |
+
"grad_norm": 16.12239646911621,
|
| 1814 |
+
"learning_rate": 2.8429125177867622e-05,
|
| 1815 |
+
"loss": 11.203057098388673,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.2962047975211586,
|
| 1820 |
+
"grad_norm": 22.192853927612305,
|
| 1821 |
+
"learning_rate": 2.8266210820616644e-05,
|
| 1822 |
+
"loss": 9.194258880615234,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.3012294760202452,
|
| 1827 |
+
"grad_norm": 16.881107330322266,
|
| 1828 |
+
"learning_rate": 2.8103155228798855e-05,
|
| 1829 |
+
"loss": 9.5519775390625,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.306254154519332,
|
| 1834 |
+
"grad_norm": 36.23583221435547,
|
| 1835 |
+
"learning_rate": 2.7939965453118318e-05,
|
| 1836 |
+
"loss": 9.040764617919923,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.3112788330184186,
|
| 1841 |
+
"grad_norm": 17.388961791992188,
|
| 1842 |
+
"learning_rate": 2.777664855008133e-05,
|
| 1843 |
+
"loss": 9.911444854736327,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.3163035115175052,
|
| 1848 |
+
"grad_norm": 109.7496566772461,
|
| 1849 |
+
"learning_rate": 2.761321158169134e-05,
|
| 1850 |
+
"loss": 9.183967590332031,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.3213281900165919,
|
| 1855 |
+
"grad_norm": 19.465408325195312,
|
| 1856 |
+
"learning_rate": 2.7449661615143545e-05,
|
| 1857 |
+
"loss": 9.167644500732422,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.3263528685156785,
|
| 1862 |
+
"grad_norm": 23.866313934326172,
|
| 1863 |
+
"learning_rate": 2.7286005722519314e-05,
|
| 1864 |
+
"loss": 10.72784652709961,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.3313775470147653,
|
| 1869 |
+
"grad_norm": 16.75422477722168,
|
| 1870 |
+
"learning_rate": 2.712225098048037e-05,
|
| 1871 |
+
"loss": 8.88067855834961,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.336402225513852,
|
| 1876 |
+
"grad_norm": 21.11101531982422,
|
| 1877 |
+
"learning_rate": 2.6958404469962794e-05,
|
| 1878 |
+
"loss": 9.210741424560547,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.3414269040129385,
|
| 1883 |
+
"grad_norm": 18.632394790649414,
|
| 1884 |
+
"learning_rate": 2.6794473275870847e-05,
|
| 1885 |
+
"loss": 8.116027069091796,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.3464515825120251,
|
| 1890 |
+
"grad_norm": 16.196258544921875,
|
| 1891 |
+
"learning_rate": 2.6630464486770584e-05,
|
| 1892 |
+
"loss": 8.565786743164063,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.3514762610111117,
|
| 1897 |
+
"grad_norm": 15.409602165222168,
|
| 1898 |
+
"learning_rate": 2.6466385194583367e-05,
|
| 1899 |
+
"loss": 7.9316253662109375,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.3565009395101986,
|
| 1904 |
+
"grad_norm": 19.422887802124023,
|
| 1905 |
+
"learning_rate": 2.6302242494279194e-05,
|
| 1906 |
+
"loss": 7.932023620605468,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.3615256180092852,
|
| 1911 |
+
"grad_norm": 17.550283432006836,
|
| 1912 |
+
"learning_rate": 2.613804348356988e-05,
|
| 1913 |
+
"loss": 8.40458984375,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.3665502965083718,
|
| 1918 |
+
"grad_norm": 21.085948944091797,
|
| 1919 |
+
"learning_rate": 2.597379526260219e-05,
|
| 1920 |
+
"loss": 7.8362586975097654,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.3715749750074586,
|
| 1925 |
+
"grad_norm": 16.09673309326172,
|
| 1926 |
+
"learning_rate": 2.5809504933650775e-05,
|
| 1927 |
+
"loss": 7.722866058349609,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.3765996535065452,
|
| 1932 |
+
"grad_norm": 13.360176086425781,
|
| 1933 |
+
"learning_rate": 2.564517960081108e-05,
|
| 1934 |
+
"loss": 8.046268463134766,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.3816243320056318,
|
| 1939 |
+
"grad_norm": 17.158672332763672,
|
| 1940 |
+
"learning_rate": 2.5480826369692178e-05,
|
| 1941 |
+
"loss": 8.267652893066407,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.3866490105047184,
|
| 1946 |
+
"grad_norm": 16.403295516967773,
|
| 1947 |
+
"learning_rate": 2.5316452347109486e-05,
|
| 1948 |
+
"loss": 7.891246795654297,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.391673689003805,
|
| 1953 |
+
"grad_norm": 18.296703338623047,
|
| 1954 |
+
"learning_rate": 2.5152064640777444e-05,
|
| 1955 |
+
"loss": 8.161449432373047,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.396698367502892,
|
| 1960 |
+
"grad_norm": 13.22531509399414,
|
| 1961 |
+
"learning_rate": 2.4987670359002225e-05,
|
| 1962 |
+
"loss": 7.458624267578125,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.4017230460019785,
|
| 1967 |
+
"grad_norm": 15.091835975646973,
|
| 1968 |
+
"learning_rate": 2.482327661037431e-05,
|
| 1969 |
+
"loss": 6.726652526855469,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.406747724501065,
|
| 1974 |
+
"grad_norm": 24.141721725463867,
|
| 1975 |
+
"learning_rate": 2.4658890503461142e-05,
|
| 1976 |
+
"loss": 7.5736839294433596,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.4117724030001517,
|
| 1981 |
+
"grad_norm": 15.211458206176758,
|
| 1982 |
+
"learning_rate": 2.4494519146499718e-05,
|
| 1983 |
+
"loss": 6.948004150390625,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.4167970814992383,
|
| 1988 |
+
"grad_norm": 18.808673858642578,
|
| 1989 |
+
"learning_rate": 2.4330169647089222e-05,
|
| 1990 |
+
"loss": 7.693637084960938,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.4218217599983252,
|
| 1995 |
+
"grad_norm": 12.74593734741211,
|
| 1996 |
+
"learning_rate": 2.416584911188371e-05,
|
| 1997 |
+
"loss": 6.476108551025391,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.4268464384974118,
|
| 2002 |
+
"grad_norm": 14.523116111755371,
|
| 2003 |
+
"learning_rate": 2.400156464628479e-05,
|
| 2004 |
+
"loss": 6.905941772460937,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.4318711169964984,
|
| 2009 |
+
"grad_norm": 13.235814094543457,
|
| 2010 |
+
"learning_rate": 2.3837323354134374e-05,
|
| 2011 |
+
"loss": 6.449698638916016,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.4368957954955852,
|
| 2016 |
+
"grad_norm": 12.135284423828125,
|
| 2017 |
+
"learning_rate": 2.3673132337407514e-05,
|
| 2018 |
+
"loss": 6.168265151977539,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.4419204739946718,
|
| 2023 |
+
"grad_norm": 14.88333511352539,
|
| 2024 |
+
"learning_rate": 2.350899869590529e-05,
|
| 2025 |
+
"loss": 6.634694671630859,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.4469451524937584,
|
| 2030 |
+
"grad_norm": 14.54021167755127,
|
| 2031 |
+
"learning_rate": 2.3344929526947816e-05,
|
| 2032 |
+
"loss": 6.686222076416016,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 1.451969830992845,
|
| 2037 |
+
"grad_norm": 14.360895156860352,
|
| 2038 |
+
"learning_rate": 2.3180931925067337e-05,
|
| 2039 |
+
"loss": 5.889413833618164,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.4569945094919317,
|
| 2044 |
+
"grad_norm": 21.827463150024414,
|
| 2045 |
+
"learning_rate": 2.3017012981701482e-05,
|
| 2046 |
+
"loss": 6.283362960815429,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.4620191879910185,
|
| 2051 |
+
"grad_norm": 14.690330505371094,
|
| 2052 |
+
"learning_rate": 2.2853179784886566e-05,
|
| 2053 |
+
"loss": 5.913167190551758,
|
| 2054 |
+
"step": 5820
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.467043866490105,
|
| 2058 |
+
"grad_norm": 14.252549171447754,
|
| 2059 |
+
"learning_rate": 2.2689439418951152e-05,
|
| 2060 |
+
"loss": 6.030026245117187,
|
| 2061 |
+
"step": 5840
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.4720685449891917,
|
| 2065 |
+
"grad_norm": 11.477004051208496,
|
| 2066 |
+
"learning_rate": 2.252579896420968e-05,
|
| 2067 |
+
"loss": 6.0997673034667965,
|
| 2068 |
+
"step": 5860
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 1.4770932234882783,
|
| 2072 |
+
"grad_norm": 13.692665100097656,
|
| 2073 |
+
"learning_rate": 2.2362265496656314e-05,
|
| 2074 |
+
"loss": 5.615577316284179,
|
| 2075 |
+
"step": 5880
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 1.482117901987365,
|
| 2079 |
+
"grad_norm": 13.419648170471191,
|
| 2080 |
+
"learning_rate": 2.2198846087659e-05,
|
| 2081 |
+
"loss": 6.581723022460937,
|
| 2082 |
+
"step": 5900
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 1.4871425804864518,
|
| 2086 |
+
"grad_norm": 13.096921920776367,
|
| 2087 |
+
"learning_rate": 2.203554780365365e-05,
|
| 2088 |
+
"loss": 6.183859252929688,
|
| 2089 |
+
"step": 5920
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 1.4921672589855384,
|
| 2093 |
+
"grad_norm": 15.762218475341797,
|
| 2094 |
+
"learning_rate": 2.187237770583859e-05,
|
| 2095 |
+
"loss": 5.984473419189453,
|
| 2096 |
+
"step": 5940
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 1.497191937484625,
|
| 2100 |
+
"grad_norm": 14.980542182922363,
|
| 2101 |
+
"learning_rate": 2.170934284986924e-05,
|
| 2102 |
+
"loss": 5.577951431274414,
|
| 2103 |
+
"step": 5960
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.5022166159837118,
|
| 2107 |
+
"grad_norm": 18.4476318359375,
|
| 2108 |
+
"learning_rate": 2.1546450285553012e-05,
|
| 2109 |
+
"loss": 5.640556716918946,
|
| 2110 |
+
"step": 5980
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 1.5072412944827982,
|
| 2114 |
+
"grad_norm": 11.552315711975098,
|
| 2115 |
+
"learning_rate": 2.1383707056544476e-05,
|
| 2116 |
+
"loss": 5.833734130859375,
|
| 2117 |
+
"step": 6000
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.512265972981885,
|
| 2121 |
+
"grad_norm": 11.801403045654297,
|
| 2122 |
+
"learning_rate": 2.122112020004078e-05,
|
| 2123 |
+
"loss": 5.025530624389648,
|
| 2124 |
+
"step": 6020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 1.5172906514809716,
|
| 2128 |
+
"grad_norm": 13.236927032470703,
|
| 2129 |
+
"learning_rate": 2.1058696746477337e-05,
|
| 2130 |
+
"loss": 5.505057907104492,
|
| 2131 |
+
"step": 6040
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 1.5223153299800583,
|
| 2135 |
+
"grad_norm": 11.760169982910156,
|
| 2136 |
+
"learning_rate": 2.0896443719223834e-05,
|
| 2137 |
+
"loss": 5.180996322631836,
|
| 2138 |
+
"step": 6060
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 1.527340008479145,
|
| 2142 |
+
"grad_norm": 13.339624404907227,
|
| 2143 |
+
"learning_rate": 2.0734368134280572e-05,
|
| 2144 |
+
"loss": 5.530018615722656,
|
| 2145 |
+
"step": 6080
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.5323646869782315,
|
| 2149 |
+
"grad_norm": 15.223504066467285,
|
| 2150 |
+
"learning_rate": 2.057247699997501e-05,
|
| 2151 |
+
"loss": 5.101480865478516,
|
| 2152 |
+
"step": 6100
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 1.5373893654773183,
|
| 2156 |
+
"grad_norm": 11.12869644165039,
|
| 2157 |
+
"learning_rate": 2.0410777316658776e-05,
|
| 2158 |
+
"loss": 5.387827301025391,
|
| 2159 |
+
"step": 6120
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 1.542414043976405,
|
| 2163 |
+
"grad_norm": 11.168490409851074,
|
| 2164 |
+
"learning_rate": 2.0249276076404937e-05,
|
| 2165 |
+
"loss": 5.365410232543946,
|
| 2166 |
+
"step": 6140
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 1.5474387224754915,
|
| 2170 |
+
"grad_norm": 11.577016830444336,
|
| 2171 |
+
"learning_rate": 2.0087980262705663e-05,
|
| 2172 |
+
"loss": 5.477450561523438,
|
| 2173 |
+
"step": 6160
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 1.5524634009745784,
|
| 2177 |
+
"grad_norm": 14.104944229125977,
|
| 2178 |
+
"learning_rate": 1.992689685017028e-05,
|
| 2179 |
+
"loss": 4.753940200805664,
|
| 2180 |
+
"step": 6180
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 1.557488079473665,
|
| 2184 |
+
"grad_norm": 10.449115753173828,
|
| 2185 |
+
"learning_rate": 1.9766032804223612e-05,
|
| 2186 |
+
"loss": 4.47644157409668,
|
| 2187 |
+
"step": 6200
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.5625127579727516,
|
| 2191 |
+
"grad_norm": 14.98228931427002,
|
| 2192 |
+
"learning_rate": 1.960539508080485e-05,
|
| 2193 |
+
"loss": 4.596777725219726,
|
| 2194 |
+
"step": 6220
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 1.5675374364718384,
|
| 2198 |
+
"grad_norm": 16.68485450744629,
|
| 2199 |
+
"learning_rate": 1.944499062606675e-05,
|
| 2200 |
+
"loss": 4.914170455932617,
|
| 2201 |
+
"step": 6240
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 1.5725621149709248,
|
| 2205 |
+
"grad_norm": 14.826934814453125,
|
| 2206 |
+
"learning_rate": 1.9284826376075256e-05,
|
| 2207 |
+
"loss": 4.500260925292968,
|
| 2208 |
+
"step": 6260
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.5775867934700116,
|
| 2212 |
+
"grad_norm": 15.89423656463623,
|
| 2213 |
+
"learning_rate": 1.9124909256509622e-05,
|
| 2214 |
+
"loss": 4.9693656921386715,
|
| 2215 |
+
"step": 6280
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 1.5826114719690982,
|
| 2219 |
+
"grad_norm": 25.843351364135742,
|
| 2220 |
+
"learning_rate": 1.896524618236289e-05,
|
| 2221 |
+
"loss": 4.7220813751220705,
|
| 2222 |
+
"step": 6300
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.6456694974026145,
|
| 2226 |
+
"grad_norm": 15.981147766113281,
|
| 2227 |
+
"learning_rate": 4.999985708624277e-05,
|
| 2228 |
+
"loss": 6.141958999633789,
|
| 2229 |
+
"step": 6320
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.6540439396516753,
|
| 2233 |
+
"grad_norm": 12.432202339172363,
|
| 2234 |
+
"learning_rate": 4.999851618030373e-05,
|
| 2235 |
+
"loss": 6.936581420898437,
|
| 2236 |
+
"step": 6340
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.6624183819007365,
|
| 2240 |
+
"grad_norm": 14.735859870910645,
|
| 2241 |
+
"learning_rate": 4.9995763869538124e-05,
|
| 2242 |
+
"loss": 7.491865539550782,
|
| 2243 |
+
"step": 6360
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.670792824149798,
|
| 2247 |
+
"grad_norm": 12.628633499145508,
|
| 2248 |
+
"learning_rate": 4.999160030934015e-05,
|
| 2249 |
+
"loss": 7.847127532958984,
|
| 2250 |
+
"step": 6380
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.679167266398859,
|
| 2254 |
+
"grad_norm": 14.50905990600586,
|
| 2255 |
+
"learning_rate": 4.9986025734782515e-05,
|
| 2256 |
+
"loss": 6.550693511962891,
|
| 2257 |
+
"step": 6400
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.6875417086479203,
|
| 2261 |
+
"grad_norm": 18.310335159301758,
|
| 2262 |
+
"learning_rate": 4.997904046060316e-05,
|
| 2263 |
+
"loss": 6.068325042724609,
|
| 2264 |
+
"step": 6420
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.695916150896981,
|
| 2268 |
+
"grad_norm": 12.4039888381958,
|
| 2269 |
+
"learning_rate": 4.997064488118748e-05,
|
| 2270 |
+
"loss": 5.966919708251953,
|
| 2271 |
+
"step": 6440
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.7042905931460424,
|
| 2275 |
+
"grad_norm": 14.189772605895996,
|
| 2276 |
+
"learning_rate": 4.996083947054607e-05,
|
| 2277 |
+
"loss": 5.816884613037109,
|
| 2278 |
+
"step": 6460
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.7126650353951036,
|
| 2282 |
+
"grad_norm": 11.678384780883789,
|
| 2283 |
+
"learning_rate": 4.9949624782287944e-05,
|
| 2284 |
+
"loss": 6.372280120849609,
|
| 2285 |
+
"step": 6480
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.721039477644165,
|
| 2289 |
+
"grad_norm": 13.517382621765137,
|
| 2290 |
+
"learning_rate": 4.993700144958927e-05,
|
| 2291 |
+
"loss": 5.818735504150391,
|
| 2292 |
+
"step": 6500
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.7294139198932257,
|
| 2296 |
+
"grad_norm": 11.332883834838867,
|
| 2297 |
+
"learning_rate": 4.992297018515767e-05,
|
| 2298 |
+
"loss": 5.053961181640625,
|
| 2299 |
+
"step": 6520
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.737788362142287,
|
| 2303 |
+
"grad_norm": 29.601367950439453,
|
| 2304 |
+
"learning_rate": 4.990753178119193e-05,
|
| 2305 |
+
"loss": 5.2946617126464846,
|
| 2306 |
+
"step": 6540
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.746162804391348,
|
| 2310 |
+
"grad_norm": 10.445046424865723,
|
| 2311 |
+
"learning_rate": 4.989068710933732e-05,
|
| 2312 |
+
"loss": 4.721809005737304,
|
| 2313 |
+
"step": 6560
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.7545372466404094,
|
| 2317 |
+
"grad_norm": 14.08784294128418,
|
| 2318 |
+
"learning_rate": 4.9872437120636304e-05,
|
| 2319 |
+
"loss": 4.68652572631836,
|
| 2320 |
+
"step": 6580
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.7629116888894707,
|
| 2324 |
+
"grad_norm": 10.382508277893066,
|
| 2325 |
+
"learning_rate": 4.985278284547495e-05,
|
| 2326 |
+
"loss": 4.90959358215332,
|
| 2327 |
+
"step": 6600
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.7712861311385315,
|
| 2331 |
+
"grad_norm": 9.340018272399902,
|
| 2332 |
+
"learning_rate": 4.983172539352465e-05,
|
| 2333 |
+
"loss": 4.94355354309082,
|
| 2334 |
+
"step": 6620
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.7796605733875928,
|
| 2338 |
+
"grad_norm": 11.943198204040527,
|
| 2339 |
+
"learning_rate": 4.980926595367956e-05,
|
| 2340 |
+
"loss": 4.442295074462891,
|
| 2341 |
+
"step": 6640
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.788035015636654,
|
| 2345 |
+
"grad_norm": 12.623554229736328,
|
| 2346 |
+
"learning_rate": 4.978540579398941e-05,
|
| 2347 |
+
"loss": 4.63629150390625,
|
| 2348 |
+
"step": 6660
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.796409457885715,
|
| 2352 |
+
"grad_norm": 9.113373756408691,
|
| 2353 |
+
"learning_rate": 4.9760146261587936e-05,
|
| 2354 |
+
"loss": 4.220952987670898,
|
| 2355 |
+
"step": 6680
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.804783900134776,
|
| 2359 |
+
"grad_norm": 9.733516693115234,
|
| 2360 |
+
"learning_rate": 4.973348878261681e-05,
|
| 2361 |
+
"loss": 4.143902206420899,
|
| 2362 |
+
"step": 6700
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.8131583423838373,
|
| 2366 |
+
"grad_norm": 10.420555114746094,
|
| 2367 |
+
"learning_rate": 4.970543486214516e-05,
|
| 2368 |
+
"loss": 4.141384506225586,
|
| 2369 |
+
"step": 6720
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.8215327846328986,
|
| 2373 |
+
"grad_norm": 7.3338212966918945,
|
| 2374 |
+
"learning_rate": 4.967598608408454e-05,
|
| 2375 |
+
"loss": 3.72052001953125,
|
| 2376 |
+
"step": 6740
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 2.82990722688196,
|
| 2380 |
+
"grad_norm": 10.468330383300781,
|
| 2381 |
+
"learning_rate": 4.9645144111099515e-05,
|
| 2382 |
+
"loss": 3.4560214996337892,
|
| 2383 |
+
"step": 6760
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 2.8382816691310206,
|
| 2387 |
+
"grad_norm": 10.967416763305664,
|
| 2388 |
+
"learning_rate": 4.961291068451385e-05,
|
| 2389 |
+
"loss": 3.2705413818359377,
|
| 2390 |
+
"step": 6780
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.846656111380082,
|
| 2394 |
+
"grad_norm": 12.951521873474121,
|
| 2395 |
+
"learning_rate": 4.9579287624212093e-05,
|
| 2396 |
+
"loss": 3.3012428283691406,
|
| 2397 |
+
"step": 6800
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.855030553629143,
|
| 2401 |
+
"grad_norm": 9.387336730957031,
|
| 2402 |
+
"learning_rate": 4.9544276828536906e-05,
|
| 2403 |
+
"loss": 3.2594974517822264,
|
| 2404 |
+
"step": 6820
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 2.863404995878204,
|
| 2408 |
+
"grad_norm": 8.693846702575684,
|
| 2409 |
+
"learning_rate": 4.9507880274181854e-05,
|
| 2410 |
+
"loss": 2.9371822357177733,
|
| 2411 |
+
"step": 6840
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.871779438127265,
|
| 2415 |
+
"grad_norm": 11.925010681152344,
|
| 2416 |
+
"learning_rate": 4.94701000160798e-05,
|
| 2417 |
+
"loss": 2.991471290588379,
|
| 2418 |
+
"step": 6860
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 2.8801538803763265,
|
| 2422 |
+
"grad_norm": 9.22618579864502,
|
| 2423 |
+
"learning_rate": 4.943093818728689e-05,
|
| 2424 |
+
"loss": 3.6338191986083985,
|
| 2425 |
+
"step": 6880
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 2.8885283226253877,
|
| 2429 |
+
"grad_norm": 9.440751075744629,
|
| 2430 |
+
"learning_rate": 4.939039699886213e-05,
|
| 2431 |
+
"loss": 2.745065689086914,
|
| 2432 |
+
"step": 6900
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 2.896902764874449,
|
| 2436 |
+
"grad_norm": 12.722249984741211,
|
| 2437 |
+
"learning_rate": 4.9348478739742526e-05,
|
| 2438 |
+
"loss": 2.652340126037598,
|
| 2439 |
+
"step": 6920
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.90527720712351,
|
| 2443 |
+
"grad_norm": 10.900734901428223,
|
| 2444 |
+
"learning_rate": 4.930518577661388e-05,
|
| 2445 |
+
"loss": 2.8322593688964846,
|
| 2446 |
+
"step": 6940
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 2.913651649372571,
|
| 2450 |
+
"grad_norm": 6.965799808502197,
|
| 2451 |
+
"learning_rate": 4.926052055377716e-05,
|
| 2452 |
+
"loss": 2.7872573852539064,
|
| 2453 |
+
"step": 6960
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.9220260916216323,
|
| 2457 |
+
"grad_norm": 6.74927282333374,
|
| 2458 |
+
"learning_rate": 4.921448559301048e-05,
|
| 2459 |
+
"loss": 2.9263822555541994,
|
| 2460 |
+
"step": 6980
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 2.930400533870693,
|
| 2464 |
+
"grad_norm": 8.862627029418945,
|
| 2465 |
+
"learning_rate": 4.916708349342674e-05,
|
| 2466 |
+
"loss": 2.499791717529297,
|
| 2467 |
+
"step": 7000
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 2.9387749761197544,
|
| 2471 |
+
"grad_norm": 10.037351608276367,
|
| 2472 |
+
"learning_rate": 4.911831693132686e-05,
|
| 2473 |
+
"loss": 2.275549125671387,
|
| 2474 |
+
"step": 7020
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 2.9471494183688156,
|
| 2478 |
+
"grad_norm": 6.2945075035095215,
|
| 2479 |
+
"learning_rate": 4.906818866004871e-05,
|
| 2480 |
+
"loss": 2.404999542236328,
|
| 2481 |
+
"step": 7040
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.955523860617877,
|
| 2485 |
+
"grad_norm": 4.414289951324463,
|
| 2486 |
+
"learning_rate": 4.901670150981164e-05,
|
| 2487 |
+
"loss": 2.1930471420288087,
|
| 2488 |
+
"step": 7060
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 2.963898302866938,
|
| 2492 |
+
"grad_norm": 5.771992206573486,
|
| 2493 |
+
"learning_rate": 4.8963858387556665e-05,
|
| 2494 |
+
"loss": 2.3143028259277343,
|
| 2495 |
+
"step": 7080
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 2.9722727451159994,
|
| 2499 |
+
"grad_norm": 6.430806636810303,
|
| 2500 |
+
"learning_rate": 4.890966227678239e-05,
|
| 2501 |
+
"loss": 2.1435062408447267,
|
| 2502 |
+
"step": 7100
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 2.98064718736506,
|
| 2506 |
+
"grad_norm": 4.818403244018555,
|
| 2507 |
+
"learning_rate": 4.885411623737651e-05,
|
| 2508 |
+
"loss": 1.8850128173828125,
|
| 2509 |
+
"step": 7120
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 2.9890216296141214,
|
| 2513 |
+
"grad_norm": 6.249518394470215,
|
| 2514 |
+
"learning_rate": 4.87972234054431e-05,
|
| 2515 |
+
"loss": 2.0096624374389647,
|
| 2516 |
+
"step": 7140
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.9973960718631827,
|
| 2520 |
+
"grad_norm": 7.708676815032959,
|
| 2521 |
+
"learning_rate": 4.87389869931255e-05,
|
| 2522 |
+
"loss": 1.9306636810302735,
|
| 2523 |
+
"step": 7160
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 3.0054433874618898,
|
| 2527 |
+
"grad_norm": 10.92924690246582,
|
| 2528 |
+
"learning_rate": 4.8679410288425e-05,
|
| 2529 |
+
"loss": 1.60665283203125,
|
| 2530 |
+
"step": 7180
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 3.013817829710951,
|
| 2534 |
+
"grad_norm": 8.802605628967285,
|
| 2535 |
+
"learning_rate": 4.86184966550152e-05,
|
| 2536 |
+
"loss": 1.6958501815795899,
|
| 2537 |
+
"step": 7200
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 3.022192271960012,
|
| 2541 |
+
"grad_norm": 11.568385124206543,
|
| 2542 |
+
"learning_rate": 4.8556249532052045e-05,
|
| 2543 |
+
"loss": 1.7447893142700195,
|
| 2544 |
+
"step": 7220
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 3.030566714209073,
|
| 2548 |
+
"grad_norm": 4.640422344207764,
|
| 2549 |
+
"learning_rate": 4.8492672433979744e-05,
|
| 2550 |
+
"loss": 1.7216526031494142,
|
| 2551 |
+
"step": 7240
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 3.0389411564581343,
|
| 2555 |
+
"grad_norm": 5.1430816650390625,
|
| 2556 |
+
"learning_rate": 4.8427768950332264e-05,
|
| 2557 |
+
"loss": 1.9419631958007812,
|
| 2558 |
+
"step": 7260
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 3.0473155987071956,
|
| 2562 |
+
"grad_norm": 4.505365371704102,
|
| 2563 |
+
"learning_rate": 4.836154274553069e-05,
|
| 2564 |
+
"loss": 1.4452963829040528,
|
| 2565 |
+
"step": 7280
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 3.0556900409562564,
|
| 2569 |
+
"grad_norm": 4.836411952972412,
|
| 2570 |
+
"learning_rate": 4.829399755867636e-05,
|
| 2571 |
+
"loss": 1.5338068962097169,
|
| 2572 |
+
"step": 7300
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 3.0640644832053177,
|
| 2576 |
+
"grad_norm": 6.093108177185059,
|
| 2577 |
+
"learning_rate": 4.8225137203339734e-05,
|
| 2578 |
+
"loss": 1.553403663635254,
|
| 2579 |
+
"step": 7320
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 3.072438925454379,
|
| 2583 |
+
"grad_norm": 6.984614372253418,
|
| 2584 |
+
"learning_rate": 4.8154965567345073e-05,
|
| 2585 |
+
"loss": 1.5232481956481934,
|
| 2586 |
+
"step": 7340
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 3.08081336770344,
|
| 2590 |
+
"grad_norm": 5.966179370880127,
|
| 2591 |
+
"learning_rate": 4.8083486612550954e-05,
|
| 2592 |
+
"loss": 1.9450897216796874,
|
| 2593 |
+
"step": 7360
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 3.0891878099525014,
|
| 2597 |
+
"grad_norm": 8.660058975219727,
|
| 2598 |
+
"learning_rate": 4.801070437462659e-05,
|
| 2599 |
+
"loss": 1.4149410247802734,
|
| 2600 |
+
"step": 7380
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 3.0975622522015622,
|
| 2604 |
+
"grad_norm": 4.126588821411133,
|
| 2605 |
+
"learning_rate": 4.793662296282393e-05,
|
| 2606 |
+
"loss": 1.0894225120544434,
|
| 2607 |
+
"step": 7400
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 3.1059366944506235,
|
| 2611 |
+
"grad_norm": 9.536477088928223,
|
| 2612 |
+
"learning_rate": 4.7861246559745725e-05,
|
| 2613 |
+
"loss": 1.519902515411377,
|
| 2614 |
+
"step": 7420
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 3.1143111366996847,
|
| 2618 |
+
"grad_norm": 5.378171443939209,
|
| 2619 |
+
"learning_rate": 4.7784579421109334e-05,
|
| 2620 |
+
"loss": 1.1999958038330079,
|
| 2621 |
+
"step": 7440
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 3.122685578948746,
|
| 2625 |
+
"grad_norm": 4.443706512451172,
|
| 2626 |
+
"learning_rate": 4.770662587550643e-05,
|
| 2627 |
+
"loss": 1.2191826820373535,
|
| 2628 |
+
"step": 7460
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 3.131060021197807,
|
| 2632 |
+
"grad_norm": 4.879043102264404,
|
| 2633 |
+
"learning_rate": 4.762739032415868e-05,
|
| 2634 |
+
"loss": 1.2340889930725099,
|
| 2635 |
+
"step": 7480
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 3.139434463446868,
|
| 2639 |
+
"grad_norm": 4.470728874206543,
|
| 2640 |
+
"learning_rate": 4.754687724066917e-05,
|
| 2641 |
+
"loss": 1.423619270324707,
|
| 2642 |
+
"step": 7500
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 3.1478089056959293,
|
| 2646 |
+
"grad_norm": 5.47915506362915,
|
| 2647 |
+
"learning_rate": 4.746509117076991e-05,
|
| 2648 |
+
"loss": 1.251408863067627,
|
| 2649 |
+
"step": 7520
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 3.1561833479449906,
|
| 2653 |
+
"grad_norm": 4.566497325897217,
|
| 2654 |
+
"learning_rate": 4.7382036732065104e-05,
|
| 2655 |
+
"loss": 1.0826839447021483,
|
| 2656 |
+
"step": 7540
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 3.1645577901940514,
|
| 2660 |
+
"grad_norm": 6.905686855316162,
|
| 2661 |
+
"learning_rate": 4.72977186137705e-05,
|
| 2662 |
+
"loss": 1.069216251373291,
|
| 2663 |
+
"step": 7560
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 3.1729322324431126,
|
| 2667 |
+
"grad_norm": 2.964932680130005,
|
| 2668 |
+
"learning_rate": 4.721214157644861e-05,
|
| 2669 |
+
"loss": 1.1391191482543945,
|
| 2670 |
+
"step": 7580
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 3.181306674692174,
|
| 2674 |
+
"grad_norm": 6.019253253936768,
|
| 2675 |
+
"learning_rate": 4.712531045173996e-05,
|
| 2676 |
+
"loss": 0.9639101028442383,
|
| 2677 |
+
"step": 7600
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 3.189681116941235,
|
| 2681 |
+
"grad_norm": 3.0169851779937744,
|
| 2682 |
+
"learning_rate": 4.703723014209025e-05,
|
| 2683 |
+
"loss": 0.934177017211914,
|
| 2684 |
+
"step": 7620
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 3.198055559190296,
|
| 2688 |
+
"grad_norm": 5.5374369621276855,
|
| 2689 |
+
"learning_rate": 4.694790562047361e-05,
|
| 2690 |
+
"loss": 0.921787166595459,
|
| 2691 |
+
"step": 7640
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 3.206430001439357,
|
| 2695 |
+
"grad_norm": 3.7313361167907715,
|
| 2696 |
+
"learning_rate": 4.685734193011182e-05,
|
| 2697 |
+
"loss": 0.9222763061523438,
|
| 2698 |
+
"step": 7660
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 3.2148044436884184,
|
| 2702 |
+
"grad_norm": 3.172354221343994,
|
| 2703 |
+
"learning_rate": 4.676554418418953e-05,
|
| 2704 |
+
"loss": 0.9454127311706543,
|
| 2705 |
+
"step": 7680
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 3.2231788859374797,
|
| 2709 |
+
"grad_norm": 2.3172996044158936,
|
| 2710 |
+
"learning_rate": 4.667251756556563e-05,
|
| 2711 |
+
"loss": 1.050184154510498,
|
| 2712 |
+
"step": 7700
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 3.2315533281865405,
|
| 2716 |
+
"grad_norm": 5.589139938354492,
|
| 2717 |
+
"learning_rate": 4.657826732648061e-05,
|
| 2718 |
+
"loss": 0.7619458675384522,
|
| 2719 |
+
"step": 7720
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 3.2399277704356018,
|
| 2723 |
+
"grad_norm": 4.144975662231445,
|
| 2724 |
+
"learning_rate": 4.648279878825998e-05,
|
| 2725 |
+
"loss": 0.8281337738037109,
|
| 2726 |
+
"step": 7740
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 3.248302212684663,
|
| 2730 |
+
"grad_norm": 3.205533266067505,
|
| 2731 |
+
"learning_rate": 4.63861173410139e-05,
|
| 2732 |
+
"loss": 0.8384788513183594,
|
| 2733 |
+
"step": 7760
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 3.2566766549337243,
|
| 2737 |
+
"grad_norm": 4.673986911773682,
|
| 2738 |
+
"learning_rate": 4.628822844333278e-05,
|
| 2739 |
+
"loss": 0.7749058723449707,
|
| 2740 |
+
"step": 7780
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 3.265051097182785,
|
| 2744 |
+
"grad_norm": 4.221221446990967,
|
| 2745 |
+
"learning_rate": 4.618913762197918e-05,
|
| 2746 |
+
"loss": 1.0583919525146483,
|
| 2747 |
+
"step": 7800
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 3.2734255394318463,
|
| 2751 |
+
"grad_norm": 2.1249191761016846,
|
| 2752 |
+
"learning_rate": 4.6088850471575695e-05,
|
| 2753 |
+
"loss": 0.6937400341033936,
|
| 2754 |
+
"step": 7820
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 3.2817999816809076,
|
| 2758 |
+
"grad_norm": 3.2157654762268066,
|
| 2759 |
+
"learning_rate": 4.59873726542891e-05,
|
| 2760 |
+
"loss": 0.7503646850585938,
|
| 2761 |
+
"step": 7840
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 3.290174423929969,
|
| 2765 |
+
"grad_norm": 2.9630126953125,
|
| 2766 |
+
"learning_rate": 4.58847098995107e-05,
|
| 2767 |
+
"loss": 0.7044057846069336,
|
| 2768 |
+
"step": 7860
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 3.29854886617903,
|
| 2772 |
+
"grad_norm": 2.9990007877349854,
|
| 2773 |
+
"learning_rate": 4.578086800353284e-05,
|
| 2774 |
+
"loss": 0.6530625343322753,
|
| 2775 |
+
"step": 7880
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 3.306923308428091,
|
| 2779 |
+
"grad_norm": 2.869370698928833,
|
| 2780 |
+
"learning_rate": 4.56758528292216e-05,
|
| 2781 |
+
"loss": 0.6877331733703613,
|
| 2782 |
+
"step": 7900
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 3.315297750677152,
|
| 2786 |
+
"grad_norm": 2.248493194580078,
|
| 2787 |
+
"learning_rate": 4.5569670305685866e-05,
|
| 2788 |
+
"loss": 0.7664458751678467,
|
| 2789 |
+
"step": 7920
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 3.3236721929262134,
|
| 2793 |
+
"grad_norm": 2.4392411708831787,
|
| 2794 |
+
"learning_rate": 4.546232642794249e-05,
|
| 2795 |
+
"loss": 0.6721152305603028,
|
| 2796 |
+
"step": 7940
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 3.332046635175274,
|
| 2800 |
+
"grad_norm": 1.9127728939056396,
|
| 2801 |
+
"learning_rate": 4.5353827256577894e-05,
|
| 2802 |
+
"loss": 0.6500507354736328,
|
| 2803 |
+
"step": 7960
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 3.3404210774243355,
|
| 2807 |
+
"grad_norm": 2.9150123596191406,
|
| 2808 |
+
"learning_rate": 4.52441789174058e-05,
|
| 2809 |
+
"loss": 0.6752653598785401,
|
| 2810 |
+
"step": 7980
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 3.3487955196733967,
|
| 2814 |
+
"grad_norm": 4.481291770935059,
|
| 2815 |
+
"learning_rate": 4.5133387601121476e-05,
|
| 2816 |
+
"loss": 0.6834932327270508,
|
| 2817 |
+
"step": 8000
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 3.357169961922458,
|
| 2821 |
+
"grad_norm": 1.990797519683838,
|
| 2822 |
+
"learning_rate": 4.502145956295211e-05,
|
| 2823 |
+
"loss": 0.6353804588317871,
|
| 2824 |
+
"step": 8020
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 3.3655444041715192,
|
| 2828 |
+
"grad_norm": 6.725353717803955,
|
| 2829 |
+
"learning_rate": 4.4908401122303705e-05,
|
| 2830 |
+
"loss": 0.6204021453857422,
|
| 2831 |
+
"step": 8040
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 3.37391884642058,
|
| 2835 |
+
"grad_norm": 3.649500846862793,
|
| 2836 |
+
"learning_rate": 4.4794218662404296e-05,
|
| 2837 |
+
"loss": 0.6903295516967773,
|
| 2838 |
+
"step": 8060
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 3.3822932886696413,
|
| 2842 |
+
"grad_norm": 3.0070176124572754,
|
| 2843 |
+
"learning_rate": 4.46789186299435e-05,
|
| 2844 |
+
"loss": 0.5644550800323487,
|
| 2845 |
+
"step": 8080
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 3.3906677309187025,
|
| 2849 |
+
"grad_norm": 3.847533941268921,
|
| 2850 |
+
"learning_rate": 4.4562507534708574e-05,
|
| 2851 |
+
"loss": 0.6999043464660645,
|
| 2852 |
+
"step": 8100
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 3.399042173167764,
|
| 2856 |
+
"grad_norm": 3.3568928241729736,
|
| 2857 |
+
"learning_rate": 4.444499194921691e-05,
|
| 2858 |
+
"loss": 0.5565146446228028,
|
| 2859 |
+
"step": 8120
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 3.4074166154168246,
|
| 2863 |
+
"grad_norm": 3.4787251949310303,
|
| 2864 |
+
"learning_rate": 4.4326378508344854e-05,
|
| 2865 |
+
"loss": 0.6062619686126709,
|
| 2866 |
+
"step": 8140
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 3.415791057665886,
|
| 2870 |
+
"grad_norm": 9.04430866241455,
|
| 2871 |
+
"learning_rate": 4.4206673908953235e-05,
|
| 2872 |
+
"loss": 0.4710696220397949,
|
| 2873 |
+
"step": 8160
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 3.424165499914947,
|
| 2877 |
+
"grad_norm": 2.6757805347442627,
|
| 2878 |
+
"learning_rate": 4.408588490950914e-05,
|
| 2879 |
+
"loss": 0.5070647716522216,
|
| 2880 |
+
"step": 8180
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 3.4325399421640084,
|
| 2884 |
+
"grad_norm": 8.834247589111328,
|
| 2885 |
+
"learning_rate": 4.3964018329704414e-05,
|
| 2886 |
+
"loss": 0.8732002258300782,
|
| 2887 |
+
"step": 8200
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 3.440914384413069,
|
| 2891 |
+
"grad_norm": 5.973219871520996,
|
| 2892 |
+
"learning_rate": 4.3841081050070595e-05,
|
| 2893 |
+
"loss": 0.5055295944213867,
|
| 2894 |
+
"step": 8220
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 3.4492888266621304,
|
| 2898 |
+
"grad_norm": 9.602012634277344,
|
| 2899 |
+
"learning_rate": 4.371708001159045e-05,
|
| 2900 |
+
"loss": 0.6795970439910889,
|
| 2901 |
+
"step": 8240
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 3.4576632689111917,
|
| 2905 |
+
"grad_norm": 1.5608420372009277,
|
| 2906 |
+
"learning_rate": 4.359202221530606e-05,
|
| 2907 |
+
"loss": 0.5012212753295898,
|
| 2908 |
+
"step": 8260
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 3.466037711160253,
|
| 2912 |
+
"grad_norm": 6.789797306060791,
|
| 2913 |
+
"learning_rate": 4.34659147219236e-05,
|
| 2914 |
+
"loss": 0.5281818866729736,
|
| 2915 |
+
"step": 8280
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 3.4744121534093138,
|
| 2919 |
+
"grad_norm": 4.547278881072998,
|
| 2920 |
+
"learning_rate": 4.333876465141465e-05,
|
| 2921 |
+
"loss": 0.5114842891693115,
|
| 2922 |
+
"step": 8300
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 3.482786595658375,
|
| 2926 |
+
"grad_norm": 1.368796467781067,
|
| 2927 |
+
"learning_rate": 4.321057918261423e-05,
|
| 2928 |
+
"loss": 0.40313048362731935,
|
| 2929 |
+
"step": 8320
|
| 2930 |
+
}
|
| 2931 |
+
],
|
| 2932 |
+
"logging_steps": 20,
|
| 2933 |
+
"max_steps": 8362,
|
| 2934 |
+
"num_input_tokens_seen": 0,
|
| 2935 |
+
"num_train_epochs": 4,
|
| 2936 |
+
"save_steps": 1000000000,
|
| 2937 |
+
"stateful_callbacks": {
|
| 2938 |
+
"TrainerControl": {
|
| 2939 |
+
"args": {
|
| 2940 |
+
"should_epoch_stop": false,
|
| 2941 |
+
"should_evaluate": false,
|
| 2942 |
+
"should_log": false,
|
| 2943 |
+
"should_save": true,
|
| 2944 |
+
"should_training_stop": false
|
| 2945 |
+
},
|
| 2946 |
+
"attributes": {}
|
| 2947 |
+
}
|
| 2948 |
+
},
|
| 2949 |
+
"total_flos": 1.9599911429619456e+16,
|
| 2950 |
+
"train_batch_size": 1,
|
| 2951 |
+
"trial_name": null,
|
| 2952 |
+
"trial_params": null
|
| 2953 |
+
}
|