publish trace-repro-todbert-lite-ar-s123
Browse files
trace-repro-todbert-lite-ar-s123/best/config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"input_dim": 768,
|
| 3 |
+
"hidden_dim": 768,
|
| 4 |
+
"output_dim": 768,
|
| 5 |
+
"num_layers": 6,
|
| 6 |
+
"num_heads": 8,
|
| 7 |
+
"dropout": 0.1,
|
| 8 |
+
"max_turns": 64,
|
| 9 |
+
"attention_mode": "autoregressive",
|
| 10 |
+
"use_speaker_embeddings": true,
|
| 11 |
+
"num_speakers": 4,
|
| 12 |
+
"layer_norm": true,
|
| 13 |
+
"ff_dim": 3072,
|
| 14 |
+
"activation": "gelu",
|
| 15 |
+
"output_residual": false,
|
| 16 |
+
"arch": "v2",
|
| 17 |
+
"head_mode": "tied_continuous",
|
| 18 |
+
"head_transform": true
|
| 19 |
+
}
|
trace-repro-todbert-lite-ar-s123/best/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:673adc062b468360937d261bc0cfa706d057ab5e764399866e2c7972a55d7a6e
|
| 3 |
+
size 179808064
|
trace-repro-todbert-lite-ar-s123/config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"input_dim": 768,
|
| 3 |
+
"hidden_dim": 768,
|
| 4 |
+
"output_dim": 768,
|
| 5 |
+
"num_layers": 6,
|
| 6 |
+
"num_heads": 8,
|
| 7 |
+
"dropout": 0.1,
|
| 8 |
+
"max_turns": 64,
|
| 9 |
+
"attention_mode": "autoregressive",
|
| 10 |
+
"use_speaker_embeddings": true,
|
| 11 |
+
"num_speakers": 4,
|
| 12 |
+
"layer_norm": true,
|
| 13 |
+
"ff_dim": 3072,
|
| 14 |
+
"activation": "gelu",
|
| 15 |
+
"output_residual": false,
|
| 16 |
+
"arch": "v2",
|
| 17 |
+
"head_mode": "tied_continuous",
|
| 18 |
+
"head_transform": true
|
| 19 |
+
}
|
trace-repro-todbert-lite-ar-s123/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:673adc062b468360937d261bc0cfa706d057ab5e764399866e2c7972a55d7a6e
|
| 3 |
+
size 179808064
|
trace-repro-todbert-lite-ar-s123/trainlog.jsonl
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 1, "train_loss": 4.91624, "val_loss": 5.24715, "epoch_sec": 79.6}
|
| 2 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 2, "train_loss": 4.12669, "val_loss": 5.05761, "epoch_sec": 65.1}
|
| 3 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 3, "train_loss": 3.93229, "val_loss": 4.96557, "epoch_sec": 64.8}
|
| 4 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 4, "train_loss": 3.7807, "val_loss": 4.92107, "epoch_sec": 65.1}
|
| 5 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 5, "train_loss": 3.62889, "val_loss": 4.8797, "epoch_sec": 66.0}
|
| 6 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 6, "train_loss": 3.47217, "val_loss": 4.8563, "epoch_sec": 65.6}
|
| 7 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 7, "train_loss": 3.31552, "val_loss": 4.85431, "epoch_sec": 65.6}
|
| 8 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 8, "train_loss": 3.17408, "val_loss": 4.84041, "epoch_sec": 65.9}
|
| 9 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 9, "train_loss": 3.04834, "val_loss": 4.8391, "epoch_sec": 65.5}
|
| 10 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 10, "train_loss": 2.94003, "val_loss": 4.81334, "epoch_sec": 65.8}
|
| 11 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 11, "train_loss": 2.84581, "val_loss": 4.79097, "epoch_sec": 65.7}
|
| 12 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 12, "train_loss": 2.76348, "val_loss": 4.77116, "epoch_sec": 65.6}
|
| 13 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 13, "train_loss": 2.69528, "val_loss": 4.76616, "epoch_sec": 65.7}
|
| 14 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 14, "train_loss": 2.64212, "val_loss": 4.75872, "epoch_sec": 66.0}
|
| 15 |
+
{"model": "lite", "mode": "ar", "seed": 123, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 15, "train_loss": 2.60867, "val_loss": 4.73952, "epoch_sec": 65.6}
|