Update normalized W-scaling checkpoint manifest
Browse files- ws_scaling_checkpoints.json +14 -14
ws_scaling_checkpoints.json
CHANGED
|
@@ -41,42 +41,42 @@
|
|
| 41 |
"tokens_seen": 2138240000,
|
| 42 |
"filename": "qwen3_0_6b_ws_scaling_step_10000.ckpt",
|
| 43 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_10000.ckpt",
|
| 44 |
-
"hf_path": "checkpoints/
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"step": 20000,
|
| 48 |
"tokens_seen": 4276480000,
|
| 49 |
"filename": "qwen3_0_6b_ws_scaling_step_20000.ckpt",
|
| 50 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_20000.ckpt",
|
| 51 |
-
"hf_path": "checkpoints/
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"step": 40000,
|
| 55 |
"tokens_seen": 8552960000,
|
| 56 |
"filename": "qwen3_0_6b_ws_scaling_step_40000.ckpt",
|
| 57 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_40000.ckpt",
|
| 58 |
-
"hf_path": "checkpoints/
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"step": 80000,
|
| 62 |
"tokens_seen": 17105920000,
|
| 63 |
"filename": "qwen3_0_6b_ws_scaling_step_80000.ckpt",
|
| 64 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_80000.ckpt",
|
| 65 |
-
"hf_path": "checkpoints/
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"step": 160000,
|
| 69 |
"tokens_seen": 34211840000,
|
| 70 |
"filename": "qwen3_0_6b_ws_scaling_step_160000.ckpt",
|
| 71 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_160000.ckpt",
|
| 72 |
-
"hf_path": "checkpoints/
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"step": 320000,
|
| 76 |
"tokens_seen": 68423680000,
|
| 77 |
"filename": "qwen3_0_6b_ws_scaling_step_320000.ckpt",
|
| 78 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_320000.ckpt",
|
| 79 |
-
"hf_path": "checkpoints/
|
| 80 |
}
|
| 81 |
]
|
| 82 |
},
|
|
@@ -113,28 +113,28 @@
|
|
| 113 |
"tokens_seen": 3947520000,
|
| 114 |
"filename": "qwen3_1_7b_ws_scaling_step_10000.ckpt",
|
| 115 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_10000.ckpt",
|
| 116 |
-
"hf_path": "checkpoints/
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"step": 20000,
|
| 120 |
"tokens_seen": 7895040000,
|
| 121 |
"filename": "qwen3_1_7b_ws_scaling_step_20000.ckpt",
|
| 122 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_20000.ckpt",
|
| 123 |
-
"hf_path": "checkpoints/
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"step": 40000,
|
| 127 |
"tokens_seen": 15790080000,
|
| 128 |
"filename": "qwen3_1_7b_ws_scaling_step_40000.ckpt",
|
| 129 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_40000.ckpt",
|
| 130 |
-
"hf_path": "checkpoints/
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"step": 80000,
|
| 134 |
"tokens_seen": 31580160000,
|
| 135 |
"filename": "qwen3_1_7b_ws_scaling_step_80000.ckpt",
|
| 136 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_80000.ckpt",
|
| 137 |
-
"hf_path": "checkpoints/
|
| 138 |
}
|
| 139 |
]
|
| 140 |
},
|
|
@@ -171,28 +171,28 @@
|
|
| 171 |
"tokens_seen": 3947520000,
|
| 172 |
"filename": "qwen3_4b_ws_scaling_step_10000.ckpt",
|
| 173 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_10000.ckpt",
|
| 174 |
-
"hf_path": "checkpoints/
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"step": 20000,
|
| 178 |
"tokens_seen": 7895040000,
|
| 179 |
"filename": "qwen3_4b_ws_scaling_step_20000.ckpt",
|
| 180 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_20000.ckpt",
|
| 181 |
-
"hf_path": "checkpoints/
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"step": 40000,
|
| 185 |
"tokens_seen": 15790080000,
|
| 186 |
"filename": "qwen3_4b_ws_scaling_step_40000.ckpt",
|
| 187 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_40000.ckpt",
|
| 188 |
-
"hf_path": "checkpoints/
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"step": 80000,
|
| 192 |
"tokens_seen": 31580160000,
|
| 193 |
"filename": "qwen3_4b_ws_scaling_step_80000.ckpt",
|
| 194 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_80000.ckpt",
|
| 195 |
-
"hf_path": "checkpoints/
|
| 196 |
}
|
| 197 |
]
|
| 198 |
}
|
|
|
|
| 41 |
"tokens_seen": 2138240000,
|
| 42 |
"filename": "qwen3_0_6b_ws_scaling_step_10000.ckpt",
|
| 43 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_10000.ckpt",
|
| 44 |
+
"hf_path": "checkpoints/0_6B/qwen/ws_scaling/dipper_0_6B_qwen_ws_scaling_step_10000.ckpt"
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"step": 20000,
|
| 48 |
"tokens_seen": 4276480000,
|
| 49 |
"filename": "qwen3_0_6b_ws_scaling_step_20000.ckpt",
|
| 50 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_20000.ckpt",
|
| 51 |
+
"hf_path": "checkpoints/0_6B/qwen/ws_scaling/dipper_0_6B_qwen_ws_scaling_step_20000.ckpt"
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"step": 40000,
|
| 55 |
"tokens_seen": 8552960000,
|
| 56 |
"filename": "qwen3_0_6b_ws_scaling_step_40000.ckpt",
|
| 57 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_40000.ckpt",
|
| 58 |
+
"hf_path": "checkpoints/0_6B/qwen/ws_scaling/dipper_0_6B_qwen_ws_scaling_step_40000.ckpt"
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"step": 80000,
|
| 62 |
"tokens_seen": 17105920000,
|
| 63 |
"filename": "qwen3_0_6b_ws_scaling_step_80000.ckpt",
|
| 64 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_80000.ckpt",
|
| 65 |
+
"hf_path": "checkpoints/0_6B/qwen/ws_scaling/dipper_0_6B_qwen_ws_scaling_step_80000.ckpt"
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"step": 160000,
|
| 69 |
"tokens_seen": 34211840000,
|
| 70 |
"filename": "qwen3_0_6b_ws_scaling_step_160000.ckpt",
|
| 71 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_160000.ckpt",
|
| 72 |
+
"hf_path": "checkpoints/0_6B/qwen/ws_scaling/dipper_0_6B_qwen_ws_scaling_step_160000.ckpt"
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"step": 320000,
|
| 76 |
"tokens_seen": 68423680000,
|
| 77 |
"filename": "qwen3_0_6b_ws_scaling_step_320000.ckpt",
|
| 78 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_0_6b/ws_scaling/qwen3_0_6b_ws_scaling_step_320000.ckpt",
|
| 79 |
+
"hf_path": "checkpoints/0_6B/qwen/ws_scaling/dipper_0_6B_qwen_ws_scaling_step_320000.ckpt"
|
| 80 |
}
|
| 81 |
]
|
| 82 |
},
|
|
|
|
| 113 |
"tokens_seen": 3947520000,
|
| 114 |
"filename": "qwen3_1_7b_ws_scaling_step_10000.ckpt",
|
| 115 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_10000.ckpt",
|
| 116 |
+
"hf_path": "checkpoints/1_7B/qwen/ws_scaling/dipper_1_7B_qwen_ws_scaling_step_10000.ckpt"
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"step": 20000,
|
| 120 |
"tokens_seen": 7895040000,
|
| 121 |
"filename": "qwen3_1_7b_ws_scaling_step_20000.ckpt",
|
| 122 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_20000.ckpt",
|
| 123 |
+
"hf_path": "checkpoints/1_7B/qwen/ws_scaling/dipper_1_7B_qwen_ws_scaling_step_20000.ckpt"
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"step": 40000,
|
| 127 |
"tokens_seen": 15790080000,
|
| 128 |
"filename": "qwen3_1_7b_ws_scaling_step_40000.ckpt",
|
| 129 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_40000.ckpt",
|
| 130 |
+
"hf_path": "checkpoints/1_7B/qwen/ws_scaling/dipper_1_7B_qwen_ws_scaling_step_40000.ckpt"
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"step": 80000,
|
| 134 |
"tokens_seen": 31580160000,
|
| 135 |
"filename": "qwen3_1_7b_ws_scaling_step_80000.ckpt",
|
| 136 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_1_7b/ws_scaling/qwen3_1_7b_ws_scaling_step_80000.ckpt",
|
| 137 |
+
"hf_path": "checkpoints/1_7B/qwen/ws_scaling/dipper_1_7B_qwen_ws_scaling_step_80000.ckpt"
|
| 138 |
}
|
| 139 |
]
|
| 140 |
},
|
|
|
|
| 171 |
"tokens_seen": 3947520000,
|
| 172 |
"filename": "qwen3_4b_ws_scaling_step_10000.ckpt",
|
| 173 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_10000.ckpt",
|
| 174 |
+
"hf_path": "checkpoints/4B/qwen/ws_scaling/dipper_4B_qwen_ws_scaling_step_10000.ckpt"
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"step": 20000,
|
| 178 |
"tokens_seen": 7895040000,
|
| 179 |
"filename": "qwen3_4b_ws_scaling_step_20000.ckpt",
|
| 180 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_20000.ckpt",
|
| 181 |
+
"hf_path": "checkpoints/4B/qwen/ws_scaling/dipper_4B_qwen_ws_scaling_step_20000.ckpt"
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"step": 40000,
|
| 185 |
"tokens_seen": 15790080000,
|
| 186 |
"filename": "qwen3_4b_ws_scaling_step_40000.ckpt",
|
| 187 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_40000.ckpt",
|
| 188 |
+
"hf_path": "checkpoints/4B/qwen/ws_scaling/dipper_4B_qwen_ws_scaling_step_40000.ckpt"
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"step": 80000,
|
| 192 |
"tokens_seen": 31580160000,
|
| 193 |
"filename": "qwen3_4b_ws_scaling_step_80000.ckpt",
|
| 194 |
"local_path": "/projects/u6dq/data/release_data/dipper/checkpoints/dipper_4b/ws_scaling/qwen3_4b_ws_scaling_step_80000.ckpt",
|
| 195 |
+
"hf_path": "checkpoints/4B/qwen/ws_scaling/dipper_4B_qwen_ws_scaling_step_80000.ckpt"
|
| 196 |
}
|
| 197 |
]
|
| 198 |
}
|