Commit
•
724f8dd
1
Parent(s):
b1eb9b5
Upload 7 files
Browse files
checkpoint-40000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.37.2"
|
| 6 |
+
}
|
checkpoint-40000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:699a26fa41d19bfa572e780061fb9bc863dcf6fa5c9f94ec7b7236448fe81b76
|
| 3 |
+
size 497774208
|
checkpoint-40000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88489e8883b915c0c883cd5c37a5e37f86acdc954c8fb9acd6b7abf55ff2e824
|
| 3 |
+
size 995642298
|
checkpoint-40000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af6d76d39f97dcc8f7e39f2eb0571b377af49a32b0cbc734af162f6290e6c146
|
| 3 |
+
size 14244
|
checkpoint-40000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f192ba9307540a4ed07a5fa3bc17ee20c1a5217aa8d1b017621637624a93c162
|
| 3 |
+
size 1064
|
checkpoint-40000/trainer_state.json
ADDED
|
@@ -0,0 +1,501 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.7597626604112047,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 40000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03,
|
| 13 |
+
"learning_rate": 4.942504944574767e-05,
|
| 14 |
+
"loss": 3.2985,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.07,
|
| 19 |
+
"learning_rate": 4.8850098891495336e-05,
|
| 20 |
+
"loss": 3.1395,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.1,
|
| 25 |
+
"learning_rate": 4.8275148337242996e-05,
|
| 26 |
+
"loss": 3.0915,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.14,
|
| 31 |
+
"learning_rate": 4.770019778299066e-05,
|
| 32 |
+
"loss": 3.0514,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.17,
|
| 37 |
+
"learning_rate": 4.712524722873833e-05,
|
| 38 |
+
"loss": 2.9975,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.21,
|
| 43 |
+
"learning_rate": 4.6550296674485996e-05,
|
| 44 |
+
"loss": 2.9986,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.24,
|
| 49 |
+
"learning_rate": 4.597534612023366e-05,
|
| 50 |
+
"loss": 2.9464,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.28,
|
| 55 |
+
"learning_rate": 4.540039556598133e-05,
|
| 56 |
+
"loss": 2.9419,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.31,
|
| 61 |
+
"learning_rate": 4.4825445011728997e-05,
|
| 62 |
+
"loss": 2.9467,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.34,
|
| 67 |
+
"learning_rate": 4.425049445747666e-05,
|
| 68 |
+
"loss": 2.9477,
|
| 69 |
+
"step": 5000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.38,
|
| 73 |
+
"learning_rate": 4.367554390322432e-05,
|
| 74 |
+
"loss": 2.9098,
|
| 75 |
+
"step": 5500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.41,
|
| 79 |
+
"learning_rate": 4.310059334897199e-05,
|
| 80 |
+
"loss": 2.9206,
|
| 81 |
+
"step": 6000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.45,
|
| 85 |
+
"learning_rate": 4.252564279471966e-05,
|
| 86 |
+
"loss": 2.908,
|
| 87 |
+
"step": 6500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.48,
|
| 91 |
+
"learning_rate": 4.1950692240467324e-05,
|
| 92 |
+
"loss": 2.8956,
|
| 93 |
+
"step": 7000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.52,
|
| 97 |
+
"learning_rate": 4.1375741686214984e-05,
|
| 98 |
+
"loss": 2.9077,
|
| 99 |
+
"step": 7500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.55,
|
| 103 |
+
"learning_rate": 4.080079113196265e-05,
|
| 104 |
+
"loss": 2.8966,
|
| 105 |
+
"step": 8000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.59,
|
| 109 |
+
"learning_rate": 4.022584057771032e-05,
|
| 110 |
+
"loss": 2.8607,
|
| 111 |
+
"step": 8500
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.62,
|
| 115 |
+
"learning_rate": 3.9650890023457984e-05,
|
| 116 |
+
"loss": 2.8856,
|
| 117 |
+
"step": 9000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.66,
|
| 121 |
+
"learning_rate": 3.907593946920565e-05,
|
| 122 |
+
"loss": 2.8774,
|
| 123 |
+
"step": 9500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.69,
|
| 127 |
+
"learning_rate": 3.850098891495332e-05,
|
| 128 |
+
"loss": 2.8594,
|
| 129 |
+
"step": 10000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.72,
|
| 133 |
+
"learning_rate": 3.7926038360700984e-05,
|
| 134 |
+
"loss": 2.8541,
|
| 135 |
+
"step": 10500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.76,
|
| 139 |
+
"learning_rate": 3.735108780644865e-05,
|
| 140 |
+
"loss": 2.8643,
|
| 141 |
+
"step": 11000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.79,
|
| 145 |
+
"learning_rate": 3.677613725219631e-05,
|
| 146 |
+
"loss": 2.8425,
|
| 147 |
+
"step": 11500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.83,
|
| 151 |
+
"learning_rate": 3.620118669794398e-05,
|
| 152 |
+
"loss": 2.8535,
|
| 153 |
+
"step": 12000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.86,
|
| 157 |
+
"learning_rate": 3.5626236143691645e-05,
|
| 158 |
+
"loss": 2.8506,
|
| 159 |
+
"step": 12500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.9,
|
| 163 |
+
"learning_rate": 3.505128558943931e-05,
|
| 164 |
+
"loss": 2.8514,
|
| 165 |
+
"step": 13000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.93,
|
| 169 |
+
"learning_rate": 3.447633503518697e-05,
|
| 170 |
+
"loss": 2.8401,
|
| 171 |
+
"step": 13500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.97,
|
| 175 |
+
"learning_rate": 3.390138448093464e-05,
|
| 176 |
+
"loss": 2.8293,
|
| 177 |
+
"step": 14000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.0,
|
| 181 |
+
"learning_rate": 3.3326433926682305e-05,
|
| 182 |
+
"loss": 2.8031,
|
| 183 |
+
"step": 14500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 1.03,
|
| 187 |
+
"learning_rate": 3.275148337242998e-05,
|
| 188 |
+
"loss": 2.704,
|
| 189 |
+
"step": 15000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.07,
|
| 193 |
+
"learning_rate": 3.217653281817764e-05,
|
| 194 |
+
"loss": 2.7216,
|
| 195 |
+
"step": 15500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 1.1,
|
| 199 |
+
"learning_rate": 3.1601582263925305e-05,
|
| 200 |
+
"loss": 2.6997,
|
| 201 |
+
"step": 16000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.14,
|
| 205 |
+
"learning_rate": 3.102663170967297e-05,
|
| 206 |
+
"loss": 2.6923,
|
| 207 |
+
"step": 16500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.17,
|
| 211 |
+
"learning_rate": 3.0451681155420635e-05,
|
| 212 |
+
"loss": 2.7136,
|
| 213 |
+
"step": 17000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.21,
|
| 217 |
+
"learning_rate": 2.9876730601168302e-05,
|
| 218 |
+
"loss": 2.6984,
|
| 219 |
+
"step": 17500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.24,
|
| 223 |
+
"learning_rate": 2.9301780046915966e-05,
|
| 224 |
+
"loss": 2.7133,
|
| 225 |
+
"step": 18000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.28,
|
| 229 |
+
"learning_rate": 2.8726829492663632e-05,
|
| 230 |
+
"loss": 2.6969,
|
| 231 |
+
"step": 18500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.31,
|
| 235 |
+
"learning_rate": 2.8151878938411296e-05,
|
| 236 |
+
"loss": 2.6954,
|
| 237 |
+
"step": 19000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 1.35,
|
| 241 |
+
"learning_rate": 2.7576928384158962e-05,
|
| 242 |
+
"loss": 2.7024,
|
| 243 |
+
"step": 19500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.38,
|
| 247 |
+
"learning_rate": 2.7001977829906626e-05,
|
| 248 |
+
"loss": 2.7208,
|
| 249 |
+
"step": 20000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.41,
|
| 253 |
+
"learning_rate": 2.6427027275654293e-05,
|
| 254 |
+
"loss": 2.683,
|
| 255 |
+
"step": 20500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.45,
|
| 259 |
+
"learning_rate": 2.5852076721401963e-05,
|
| 260 |
+
"loss": 2.6896,
|
| 261 |
+
"step": 21000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.48,
|
| 265 |
+
"learning_rate": 2.527712616714963e-05,
|
| 266 |
+
"loss": 2.7004,
|
| 267 |
+
"step": 21500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.52,
|
| 271 |
+
"learning_rate": 2.470217561289729e-05,
|
| 272 |
+
"loss": 2.7053,
|
| 273 |
+
"step": 22000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.55,
|
| 277 |
+
"learning_rate": 2.412722505864496e-05,
|
| 278 |
+
"loss": 2.6863,
|
| 279 |
+
"step": 22500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.59,
|
| 283 |
+
"learning_rate": 2.3552274504392623e-05,
|
| 284 |
+
"loss": 2.6896,
|
| 285 |
+
"step": 23000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.62,
|
| 289 |
+
"learning_rate": 2.297732395014029e-05,
|
| 290 |
+
"loss": 2.6897,
|
| 291 |
+
"step": 23500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.66,
|
| 295 |
+
"learning_rate": 2.2402373395887953e-05,
|
| 296 |
+
"loss": 2.6733,
|
| 297 |
+
"step": 24000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.69,
|
| 301 |
+
"learning_rate": 2.182742284163562e-05,
|
| 302 |
+
"loss": 2.684,
|
| 303 |
+
"step": 24500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.72,
|
| 307 |
+
"learning_rate": 2.1252472287383283e-05,
|
| 308 |
+
"loss": 2.6966,
|
| 309 |
+
"step": 25000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.76,
|
| 313 |
+
"learning_rate": 2.0677521733130954e-05,
|
| 314 |
+
"loss": 2.6691,
|
| 315 |
+
"step": 25500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.79,
|
| 319 |
+
"learning_rate": 2.0102571178878617e-05,
|
| 320 |
+
"loss": 2.6901,
|
| 321 |
+
"step": 26000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.83,
|
| 325 |
+
"learning_rate": 1.9527620624626284e-05,
|
| 326 |
+
"loss": 2.6852,
|
| 327 |
+
"step": 26500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.86,
|
| 331 |
+
"learning_rate": 1.8952670070373947e-05,
|
| 332 |
+
"loss": 2.679,
|
| 333 |
+
"step": 27000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.9,
|
| 337 |
+
"learning_rate": 1.8377719516121614e-05,
|
| 338 |
+
"loss": 2.6777,
|
| 339 |
+
"step": 27500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.93,
|
| 343 |
+
"learning_rate": 1.7802768961869277e-05,
|
| 344 |
+
"loss": 2.6864,
|
| 345 |
+
"step": 28000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.97,
|
| 349 |
+
"learning_rate": 1.7227818407616947e-05,
|
| 350 |
+
"loss": 2.6714,
|
| 351 |
+
"step": 28500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 2.0,
|
| 355 |
+
"learning_rate": 1.665286785336461e-05,
|
| 356 |
+
"loss": 2.681,
|
| 357 |
+
"step": 29000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 2.04,
|
| 361 |
+
"learning_rate": 1.6077917299112278e-05,
|
| 362 |
+
"loss": 2.6028,
|
| 363 |
+
"step": 29500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.07,
|
| 367 |
+
"learning_rate": 1.550296674485994e-05,
|
| 368 |
+
"loss": 2.5981,
|
| 369 |
+
"step": 30000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.1,
|
| 373 |
+
"learning_rate": 1.4928016190607608e-05,
|
| 374 |
+
"loss": 2.5976,
|
| 375 |
+
"step": 30500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.14,
|
| 379 |
+
"learning_rate": 1.4353065636355273e-05,
|
| 380 |
+
"loss": 2.5921,
|
| 381 |
+
"step": 31000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.17,
|
| 385 |
+
"learning_rate": 1.3778115082102941e-05,
|
| 386 |
+
"loss": 2.588,
|
| 387 |
+
"step": 31500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 2.21,
|
| 391 |
+
"learning_rate": 1.3203164527850606e-05,
|
| 392 |
+
"loss": 2.5888,
|
| 393 |
+
"step": 32000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 2.24,
|
| 397 |
+
"learning_rate": 1.2628213973598271e-05,
|
| 398 |
+
"loss": 2.6143,
|
| 399 |
+
"step": 32500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.28,
|
| 403 |
+
"learning_rate": 1.2053263419345937e-05,
|
| 404 |
+
"loss": 2.5874,
|
| 405 |
+
"step": 33000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.31,
|
| 409 |
+
"learning_rate": 1.1478312865093603e-05,
|
| 410 |
+
"loss": 2.5905,
|
| 411 |
+
"step": 33500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 2.35,
|
| 415 |
+
"learning_rate": 1.0903362310841268e-05,
|
| 416 |
+
"loss": 2.6078,
|
| 417 |
+
"step": 34000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 2.38,
|
| 421 |
+
"learning_rate": 1.0328411756588933e-05,
|
| 422 |
+
"loss": 2.5793,
|
| 423 |
+
"step": 34500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.41,
|
| 427 |
+
"learning_rate": 9.7534612023366e-06,
|
| 428 |
+
"loss": 2.5968,
|
| 429 |
+
"step": 35000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.45,
|
| 433 |
+
"learning_rate": 9.178510648084265e-06,
|
| 434 |
+
"loss": 2.595,
|
| 435 |
+
"step": 35500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 2.48,
|
| 439 |
+
"learning_rate": 8.60356009383193e-06,
|
| 440 |
+
"loss": 2.5673,
|
| 441 |
+
"step": 36000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.52,
|
| 445 |
+
"learning_rate": 8.028609539579597e-06,
|
| 446 |
+
"loss": 2.5954,
|
| 447 |
+
"step": 36500
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.55,
|
| 451 |
+
"learning_rate": 7.453658985327262e-06,
|
| 452 |
+
"loss": 2.5939,
|
| 453 |
+
"step": 37000
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 2.59,
|
| 457 |
+
"learning_rate": 6.878708431074927e-06,
|
| 458 |
+
"loss": 2.6008,
|
| 459 |
+
"step": 37500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.62,
|
| 463 |
+
"learning_rate": 6.303757876822594e-06,
|
| 464 |
+
"loss": 2.6016,
|
| 465 |
+
"step": 38000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 2.66,
|
| 469 |
+
"learning_rate": 5.728807322570259e-06,
|
| 470 |
+
"loss": 2.5827,
|
| 471 |
+
"step": 38500
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.69,
|
| 475 |
+
"learning_rate": 5.153856768317925e-06,
|
| 476 |
+
"loss": 2.5904,
|
| 477 |
+
"step": 39000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 2.73,
|
| 481 |
+
"learning_rate": 4.57890621406559e-06,
|
| 482 |
+
"loss": 2.5745,
|
| 483 |
+
"step": 39500
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 2.76,
|
| 487 |
+
"learning_rate": 4.003955659813256e-06,
|
| 488 |
+
"loss": 2.5777,
|
| 489 |
+
"step": 40000
|
| 490 |
+
}
|
| 491 |
+
],
|
| 492 |
+
"logging_steps": 500,
|
| 493 |
+
"max_steps": 43482,
|
| 494 |
+
"num_input_tokens_seen": 0,
|
| 495 |
+
"num_train_epochs": 3,
|
| 496 |
+
"save_steps": 10000,
|
| 497 |
+
"total_flos": 1.0451289341952e+16,
|
| 498 |
+
"train_batch_size": 4,
|
| 499 |
+
"trial_name": null,
|
| 500 |
+
"trial_params": null
|
| 501 |
+
}
|
checkpoint-40000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd804f0f82c5f2486fec921e97a71c484ee4e10b3248ae62bd498abef28cb29c
|
| 3 |
+
size 4664
|