Upload ANDREA-12M-TRAIN.json with huggingface_hub
Browse files- ANDREA-12M-TRAIN.json +45 -0
ANDREA-12M-TRAIN.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "megachat-v8-embd384-chatbot",
|
| 3 |
+
"dataset": "megachat-v8",
|
| 4 |
+
"curriculum": "firehose",
|
| 5 |
+
"config": {
|
| 6 |
+
"n_embd": 384,
|
| 7 |
+
"n_head": 12,
|
| 8 |
+
"n_layer": 6,
|
| 9 |
+
"block_size": 1024,
|
| 10 |
+
"vocab_type": "harris",
|
| 11 |
+
"segments": 2048,
|
| 12 |
+
"lr": 0.0004,
|
| 13 |
+
"steps": 60000,
|
| 14 |
+
"checkpoint_every": 100,
|
| 15 |
+
"sample_every": 100,
|
| 16 |
+
"sample_tokens": 420,
|
| 17 |
+
"lr_schedule": "cosine",
|
| 18 |
+
"batch_size": 8,
|
| 19 |
+
"exclude_sources": [
|
| 20 |
+
"real-tool-calls", "synthetic-bash", "tool-calls",
|
| 21 |
+
"hermes3-code", "hermes3-math",
|
| 22 |
+
"chat", "smoltalk", "oasst", "dolly",
|
| 23 |
+
"unfirehose-chat", "synthetic-chat",
|
| 24 |
+
"irc", "unweapon",
|
| 25 |
+
"repo-docs", "repo-docstrings", "repo-commits",
|
| 26 |
+
"gutenberg"
|
| 27 |
+
],
|
| 28 |
+
"source_floors": {
|
| 29 |
+
"hermes3-general": 0.8,
|
| 30 |
+
"hermes3-creative": 0.7,
|
| 31 |
+
"hermes3-roleplay": 0.7,
|
| 32 |
+
"dictionary": 0.7,
|
| 33 |
+
"gutenberg": 0.7
|
| 34 |
+
},
|
| 35 |
+
"bandit_focus_count": 2,
|
| 36 |
+
"bandit_dice_sides": 3
|
| 37 |
+
},
|
| 38 |
+
"_notes": {
|
| 39 |
+
"goal": "ANDREA phase 2 — focused curriculum, 5 arms, 2-eye bandit",
|
| 40 |
+
"params": "12.8M",
|
| 41 |
+
"arms": "hermes3-general, hermes3-creative, hermes3-roleplay, dictionary, gutenberg",
|
| 42 |
+
"bandit": "2 focus eyes, 1d3: 0=2random, 1=1random+1bandit, 2=2bandit",
|
| 43 |
+
"estimated_time": "remaining ~18K steps"
|
| 44 |
+
}
|
| 45 |
+
}
|