Training in progress, step 2400, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 847599616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7eb1052ffa88acc2589ec671755d9190e1eaf9306e47ddb49015761b6fa984d
|
| 3 |
size 847599616
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 350603
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aad373e568ae72c3ed658f34a51602cd2ab119dd24ac67405bf4d8c3a99c903a
|
| 3 |
size 350603
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c94cda14b00b5179dce1cc8d50f4aaf580e7cc7bdf97407ca11afa69eabdd060
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf2bdc8cc88b23d9d44b265f5871b271cd9c18ffb31f145316331e6dedee5009
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 150,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -3074,6 +3074,444 @@
|
|
| 3074 |
"eval_samples_per_second": 9.639,
|
| 3075 |
"eval_steps_per_second": 1.639,
|
| 3076 |
"step": 2100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3077 |
}
|
| 3078 |
],
|
| 3079 |
"logging_steps": 5,
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.48,
|
| 6 |
"eval_steps": 150,
|
| 7 |
+
"global_step": 2400,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 3074 |
"eval_samples_per_second": 9.639,
|
| 3075 |
"eval_steps_per_second": 1.639,
|
| 3076 |
"step": 2100
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.421,
|
| 3080 |
+
"grad_norm": 0.08158699423074722,
|
| 3081 |
+
"learning_rate": 2.675608645857815e-06,
|
| 3082 |
+
"loss": 6.372353363037109,
|
| 3083 |
+
"step": 2105
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 0.422,
|
| 3087 |
+
"grad_norm": 0.06957968324422836,
|
| 3088 |
+
"learning_rate": 2.6693798716464858e-06,
|
| 3089 |
+
"loss": 6.340359497070312,
|
| 3090 |
+
"step": 2110
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 0.423,
|
| 3094 |
+
"grad_norm": 0.08563192933797836,
|
| 3095 |
+
"learning_rate": 2.6631437772041238e-06,
|
| 3096 |
+
"loss": 6.318928909301758,
|
| 3097 |
+
"step": 2115
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 0.424,
|
| 3101 |
+
"grad_norm": 0.11714059859514236,
|
| 3102 |
+
"learning_rate": 2.656900430727659e-06,
|
| 3103 |
+
"loss": 6.435954284667969,
|
| 3104 |
+
"step": 2120
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 0.425,
|
| 3108 |
+
"grad_norm": 0.11297214776277542,
|
| 3109 |
+
"learning_rate": 2.650649900493329e-06,
|
| 3110 |
+
"loss": 6.496327972412109,
|
| 3111 |
+
"step": 2125
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 0.426,
|
| 3115 |
+
"grad_norm": 0.08787044882774353,
|
| 3116 |
+
"learning_rate": 2.6443922548559303e-06,
|
| 3117 |
+
"loss": 6.389971160888672,
|
| 3118 |
+
"step": 2130
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 0.427,
|
| 3122 |
+
"grad_norm": 0.07244838774204254,
|
| 3123 |
+
"learning_rate": 2.638127562248073e-06,
|
| 3124 |
+
"loss": 6.446420288085937,
|
| 3125 |
+
"step": 2135
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 0.428,
|
| 3129 |
+
"grad_norm": 0.06608100235462189,
|
| 3130 |
+
"learning_rate": 2.6318558911794333e-06,
|
| 3131 |
+
"loss": 6.427850341796875,
|
| 3132 |
+
"step": 2140
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.429,
|
| 3136 |
+
"grad_norm": 0.07050882279872894,
|
| 3137 |
+
"learning_rate": 2.625577310235998e-06,
|
| 3138 |
+
"loss": 6.362213516235352,
|
| 3139 |
+
"step": 2145
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 0.43,
|
| 3143 |
+
"grad_norm": 0.08506868034601212,
|
| 3144 |
+
"learning_rate": 2.619291888079326e-06,
|
| 3145 |
+
"loss": 6.363969039916992,
|
| 3146 |
+
"step": 2150
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 0.431,
|
| 3150 |
+
"grad_norm": 0.08637284487485886,
|
| 3151 |
+
"learning_rate": 2.612999693445784e-06,
|
| 3152 |
+
"loss": 6.497118377685547,
|
| 3153 |
+
"step": 2155
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 0.432,
|
| 3157 |
+
"grad_norm": 0.10498792678117752,
|
| 3158 |
+
"learning_rate": 2.606700795145807e-06,
|
| 3159 |
+
"loss": 6.395422744750976,
|
| 3160 |
+
"step": 2160
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 0.433,
|
| 3164 |
+
"grad_norm": 0.0660381093621254,
|
| 3165 |
+
"learning_rate": 2.6003952620631354e-06,
|
| 3166 |
+
"loss": 6.355922698974609,
|
| 3167 |
+
"step": 2165
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 0.434,
|
| 3171 |
+
"grad_norm": 0.1197703629732132,
|
| 3172 |
+
"learning_rate": 2.5940831631540697e-06,
|
| 3173 |
+
"loss": 6.523306274414063,
|
| 3174 |
+
"step": 2170
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 0.435,
|
| 3178 |
+
"grad_norm": 0.1967087686061859,
|
| 3179 |
+
"learning_rate": 2.5877645674467124e-06,
|
| 3180 |
+
"loss": 6.463238525390625,
|
| 3181 |
+
"step": 2175
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 0.436,
|
| 3185 |
+
"grad_norm": 0.09303022921085358,
|
| 3186 |
+
"learning_rate": 2.581439544040214e-06,
|
| 3187 |
+
"loss": 6.421237182617188,
|
| 3188 |
+
"step": 2180
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 0.437,
|
| 3192 |
+
"grad_norm": 0.09917596727609634,
|
| 3193 |
+
"learning_rate": 2.575108162104017e-06,
|
| 3194 |
+
"loss": 6.473545074462891,
|
| 3195 |
+
"step": 2185
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 0.438,
|
| 3199 |
+
"grad_norm": 0.072263203561306,
|
| 3200 |
+
"learning_rate": 2.568770490877099e-06,
|
| 3201 |
+
"loss": 6.432736206054687,
|
| 3202 |
+
"step": 2190
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 0.439,
|
| 3206 |
+
"grad_norm": 0.1333879828453064,
|
| 3207 |
+
"learning_rate": 2.5624265996672183e-06,
|
| 3208 |
+
"loss": 6.448049926757813,
|
| 3209 |
+
"step": 2195
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 0.44,
|
| 3213 |
+
"grad_norm": 0.08322051167488098,
|
| 3214 |
+
"learning_rate": 2.556076557850152e-06,
|
| 3215 |
+
"loss": 6.379525375366211,
|
| 3216 |
+
"step": 2200
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 0.441,
|
| 3220 |
+
"grad_norm": 0.09212733060121536,
|
| 3221 |
+
"learning_rate": 2.5497204348689403e-06,
|
| 3222 |
+
"loss": 6.376298141479492,
|
| 3223 |
+
"step": 2205
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 0.442,
|
| 3227 |
+
"grad_norm": 0.1000766009092331,
|
| 3228 |
+
"learning_rate": 2.5433583002331268e-06,
|
| 3229 |
+
"loss": 6.45618896484375,
|
| 3230 |
+
"step": 2210
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 0.443,
|
| 3234 |
+
"grad_norm": 0.07744766026735306,
|
| 3235 |
+
"learning_rate": 2.5369902235179948e-06,
|
| 3236 |
+
"loss": 6.418743133544922,
|
| 3237 |
+
"step": 2215
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 0.444,
|
| 3241 |
+
"grad_norm": 0.09729813784360886,
|
| 3242 |
+
"learning_rate": 2.530616274363813e-06,
|
| 3243 |
+
"loss": 6.469072723388672,
|
| 3244 |
+
"step": 2220
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 0.445,
|
| 3248 |
+
"grad_norm": 0.10913117229938507,
|
| 3249 |
+
"learning_rate": 2.5242365224750657e-06,
|
| 3250 |
+
"loss": 6.3639171600341795,
|
| 3251 |
+
"step": 2225
|
| 3252 |
+
},
|
| 3253 |
+
{
|
| 3254 |
+
"epoch": 0.446,
|
| 3255 |
+
"grad_norm": 0.07157459110021591,
|
| 3256 |
+
"learning_rate": 2.5178510376196997e-06,
|
| 3257 |
+
"loss": 6.399369430541992,
|
| 3258 |
+
"step": 2230
|
| 3259 |
+
},
|
| 3260 |
+
{
|
| 3261 |
+
"epoch": 0.447,
|
| 3262 |
+
"grad_norm": 0.07375568896532059,
|
| 3263 |
+
"learning_rate": 2.5114598896283524e-06,
|
| 3264 |
+
"loss": 6.35676383972168,
|
| 3265 |
+
"step": 2235
|
| 3266 |
+
},
|
| 3267 |
+
{
|
| 3268 |
+
"epoch": 0.448,
|
| 3269 |
+
"grad_norm": 0.09739139676094055,
|
| 3270 |
+
"learning_rate": 2.505063148393595e-06,
|
| 3271 |
+
"loss": 6.3606117248535154,
|
| 3272 |
+
"step": 2240
|
| 3273 |
+
},
|
| 3274 |
+
{
|
| 3275 |
+
"epoch": 0.449,
|
| 3276 |
+
"grad_norm": 0.08083254098892212,
|
| 3277 |
+
"learning_rate": 2.4986608838691663e-06,
|
| 3278 |
+
"loss": 6.317353439331055,
|
| 3279 |
+
"step": 2245
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 0.45,
|
| 3283 |
+
"grad_norm": 0.07497039437294006,
|
| 3284 |
+
"learning_rate": 2.4922531660692034e-06,
|
| 3285 |
+
"loss": 6.349576187133789,
|
| 3286 |
+
"step": 2250
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 0.45,
|
| 3290 |
+
"eval_accuracy": 0.14222466422466423,
|
| 3291 |
+
"eval_loss": 6.2949748039245605,
|
| 3292 |
+
"eval_runtime": 10.4132,
|
| 3293 |
+
"eval_samples_per_second": 9.603,
|
| 3294 |
+
"eval_steps_per_second": 1.633,
|
| 3295 |
+
"step": 2250
|
| 3296 |
+
},
|
| 3297 |
+
{
|
| 3298 |
+
"epoch": 0.451,
|
| 3299 |
+
"grad_norm": 0.0646035447716713,
|
| 3300 |
+
"learning_rate": 2.4858400650674826e-06,
|
| 3301 |
+
"loss": 6.4240165710449215,
|
| 3302 |
+
"step": 2255
|
| 3303 |
+
},
|
| 3304 |
+
{
|
| 3305 |
+
"epoch": 0.452,
|
| 3306 |
+
"grad_norm": 0.08820938318967819,
|
| 3307 |
+
"learning_rate": 2.4794216509966485e-06,
|
| 3308 |
+
"loss": 6.355297470092774,
|
| 3309 |
+
"step": 2260
|
| 3310 |
+
},
|
| 3311 |
+
{
|
| 3312 |
+
"epoch": 0.453,
|
| 3313 |
+
"grad_norm": 0.09034152328968048,
|
| 3314 |
+
"learning_rate": 2.4729979940474495e-06,
|
| 3315 |
+
"loss": 6.415824890136719,
|
| 3316 |
+
"step": 2265
|
| 3317 |
+
},
|
| 3318 |
+
{
|
| 3319 |
+
"epoch": 0.454,
|
| 3320 |
+
"grad_norm": 0.07817864418029785,
|
| 3321 |
+
"learning_rate": 2.4665691644679685e-06,
|
| 3322 |
+
"loss": 6.402967071533203,
|
| 3323 |
+
"step": 2270
|
| 3324 |
+
},
|
| 3325 |
+
{
|
| 3326 |
+
"epoch": 0.455,
|
| 3327 |
+
"grad_norm": 0.08539751917123795,
|
| 3328 |
+
"learning_rate": 2.460135232562856e-06,
|
| 3329 |
+
"loss": 6.40893783569336,
|
| 3330 |
+
"step": 2275
|
| 3331 |
+
},
|
| 3332 |
+
{
|
| 3333 |
+
"epoch": 0.456,
|
| 3334 |
+
"grad_norm": 0.11367590725421906,
|
| 3335 |
+
"learning_rate": 2.4536962686925595e-06,
|
| 3336 |
+
"loss": 6.441493225097656,
|
| 3337 |
+
"step": 2280
|
| 3338 |
+
},
|
| 3339 |
+
{
|
| 3340 |
+
"epoch": 0.457,
|
| 3341 |
+
"grad_norm": 0.09240754693746567,
|
| 3342 |
+
"learning_rate": 2.4472523432725577e-06,
|
| 3343 |
+
"loss": 6.361612319946289,
|
| 3344 |
+
"step": 2285
|
| 3345 |
+
},
|
| 3346 |
+
{
|
| 3347 |
+
"epoch": 0.458,
|
| 3348 |
+
"grad_norm": 0.12638594210147858,
|
| 3349 |
+
"learning_rate": 2.4408035267725862e-06,
|
| 3350 |
+
"loss": 6.414551544189453,
|
| 3351 |
+
"step": 2290
|
| 3352 |
+
},
|
| 3353 |
+
{
|
| 3354 |
+
"epoch": 0.459,
|
| 3355 |
+
"grad_norm": 0.10394217073917389,
|
| 3356 |
+
"learning_rate": 2.434349889715869e-06,
|
| 3357 |
+
"loss": 6.383968734741211,
|
| 3358 |
+
"step": 2295
|
| 3359 |
+
},
|
| 3360 |
+
{
|
| 3361 |
+
"epoch": 0.46,
|
| 3362 |
+
"grad_norm": 0.12000339478254318,
|
| 3363 |
+
"learning_rate": 2.4278915026783463e-06,
|
| 3364 |
+
"loss": 6.412042236328125,
|
| 3365 |
+
"step": 2300
|
| 3366 |
+
},
|
| 3367 |
+
{
|
| 3368 |
+
"epoch": 0.461,
|
| 3369 |
+
"grad_norm": 0.07175832241773605,
|
| 3370 |
+
"learning_rate": 2.421428436287905e-06,
|
| 3371 |
+
"loss": 6.3555866241455075,
|
| 3372 |
+
"step": 2305
|
| 3373 |
+
},
|
| 3374 |
+
{
|
| 3375 |
+
"epoch": 0.462,
|
| 3376 |
+
"grad_norm": 0.10095806419849396,
|
| 3377 |
+
"learning_rate": 2.4149607612236027e-06,
|
| 3378 |
+
"loss": 6.40416259765625,
|
| 3379 |
+
"step": 2310
|
| 3380 |
+
},
|
| 3381 |
+
{
|
| 3382 |
+
"epoch": 0.463,
|
| 3383 |
+
"grad_norm": 0.12205059826374054,
|
| 3384 |
+
"learning_rate": 2.408488548214899e-06,
|
| 3385 |
+
"loss": 6.329568862915039,
|
| 3386 |
+
"step": 2315
|
| 3387 |
+
},
|
| 3388 |
+
{
|
| 3389 |
+
"epoch": 0.464,
|
| 3390 |
+
"grad_norm": 0.06791554391384125,
|
| 3391 |
+
"learning_rate": 2.4020118680408777e-06,
|
| 3392 |
+
"loss": 6.424092102050781,
|
| 3393 |
+
"step": 2320
|
| 3394 |
+
},
|
| 3395 |
+
{
|
| 3396 |
+
"epoch": 0.465,
|
| 3397 |
+
"grad_norm": 0.08330659568309784,
|
| 3398 |
+
"learning_rate": 2.395530791529475e-06,
|
| 3399 |
+
"loss": 6.388699722290039,
|
| 3400 |
+
"step": 2325
|
| 3401 |
+
},
|
| 3402 |
+
{
|
| 3403 |
+
"epoch": 0.466,
|
| 3404 |
+
"grad_norm": 0.10129342973232269,
|
| 3405 |
+
"learning_rate": 2.389045389556706e-06,
|
| 3406 |
+
"loss": 6.319640350341797,
|
| 3407 |
+
"step": 2330
|
| 3408 |
+
},
|
| 3409 |
+
{
|
| 3410 |
+
"epoch": 0.467,
|
| 3411 |
+
"grad_norm": 0.07156874239444733,
|
| 3412 |
+
"learning_rate": 2.3825557330458873e-06,
|
| 3413 |
+
"loss": 6.517421722412109,
|
| 3414 |
+
"step": 2335
|
| 3415 |
+
},
|
| 3416 |
+
{
|
| 3417 |
+
"epoch": 0.468,
|
| 3418 |
+
"grad_norm": 0.10998792201280594,
|
| 3419 |
+
"learning_rate": 2.376061892966863e-06,
|
| 3420 |
+
"loss": 6.415800476074219,
|
| 3421 |
+
"step": 2340
|
| 3422 |
+
},
|
| 3423 |
+
{
|
| 3424 |
+
"epoch": 0.469,
|
| 3425 |
+
"grad_norm": 0.07388386875391006,
|
| 3426 |
+
"learning_rate": 2.3695639403352275e-06,
|
| 3427 |
+
"loss": 6.389354705810547,
|
| 3428 |
+
"step": 2345
|
| 3429 |
+
},
|
| 3430 |
+
{
|
| 3431 |
+
"epoch": 0.47,
|
| 3432 |
+
"grad_norm": 0.07077425718307495,
|
| 3433 |
+
"learning_rate": 2.36306194621155e-06,
|
| 3434 |
+
"loss": 6.414984130859375,
|
| 3435 |
+
"step": 2350
|
| 3436 |
+
},
|
| 3437 |
+
{
|
| 3438 |
+
"epoch": 0.471,
|
| 3439 |
+
"grad_norm": 0.1409406214952469,
|
| 3440 |
+
"learning_rate": 2.3565559817005965e-06,
|
| 3441 |
+
"loss": 6.4110252380371096,
|
| 3442 |
+
"step": 2355
|
| 3443 |
+
},
|
| 3444 |
+
{
|
| 3445 |
+
"epoch": 0.472,
|
| 3446 |
+
"grad_norm": 0.12504221498966217,
|
| 3447 |
+
"learning_rate": 2.350046117950552e-06,
|
| 3448 |
+
"loss": 6.524173736572266,
|
| 3449 |
+
"step": 2360
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 0.473,
|
| 3453 |
+
"grad_norm": 0.09057491272687912,
|
| 3454 |
+
"learning_rate": 2.3435324261522446e-06,
|
| 3455 |
+
"loss": 6.359022521972657,
|
| 3456 |
+
"step": 2365
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 0.474,
|
| 3460 |
+
"grad_norm": 0.08076512813568115,
|
| 3461 |
+
"learning_rate": 2.337014977538363e-06,
|
| 3462 |
+
"loss": 6.364510345458984,
|
| 3463 |
+
"step": 2370
|
| 3464 |
+
},
|
| 3465 |
+
{
|
| 3466 |
+
"epoch": 0.475,
|
| 3467 |
+
"grad_norm": 0.07883577793836594,
|
| 3468 |
+
"learning_rate": 2.330493843382682e-06,
|
| 3469 |
+
"loss": 6.344949340820312,
|
| 3470 |
+
"step": 2375
|
| 3471 |
+
},
|
| 3472 |
+
{
|
| 3473 |
+
"epoch": 0.476,
|
| 3474 |
+
"grad_norm": 0.07244715094566345,
|
| 3475 |
+
"learning_rate": 2.323969094999279e-06,
|
| 3476 |
+
"loss": 6.497917938232422,
|
| 3477 |
+
"step": 2380
|
| 3478 |
+
},
|
| 3479 |
+
{
|
| 3480 |
+
"epoch": 0.477,
|
| 3481 |
+
"grad_norm": 0.06551758199930191,
|
| 3482 |
+
"learning_rate": 2.3174408037417577e-06,
|
| 3483 |
+
"loss": 6.289946746826172,
|
| 3484 |
+
"step": 2385
|
| 3485 |
+
},
|
| 3486 |
+
{
|
| 3487 |
+
"epoch": 0.478,
|
| 3488 |
+
"grad_norm": 0.08715993165969849,
|
| 3489 |
+
"learning_rate": 2.310909041002466e-06,
|
| 3490 |
+
"loss": 6.448321533203125,
|
| 3491 |
+
"step": 2390
|
| 3492 |
+
},
|
| 3493 |
+
{
|
| 3494 |
+
"epoch": 0.479,
|
| 3495 |
+
"grad_norm": 0.08652404695749283,
|
| 3496 |
+
"learning_rate": 2.3043738782117127e-06,
|
| 3497 |
+
"loss": 6.329542922973633,
|
| 3498 |
+
"step": 2395
|
| 3499 |
+
},
|
| 3500 |
+
{
|
| 3501 |
+
"epoch": 0.48,
|
| 3502 |
+
"grad_norm": 0.06743095815181732,
|
| 3503 |
+
"learning_rate": 2.297835386836993e-06,
|
| 3504 |
+
"loss": 6.365715026855469,
|
| 3505 |
+
"step": 2400
|
| 3506 |
+
},
|
| 3507 |
+
{
|
| 3508 |
+
"epoch": 0.48,
|
| 3509 |
+
"eval_accuracy": 0.1423003663003663,
|
| 3510 |
+
"eval_loss": 6.278029441833496,
|
| 3511 |
+
"eval_runtime": 10.2925,
|
| 3512 |
+
"eval_samples_per_second": 9.716,
|
| 3513 |
+
"eval_steps_per_second": 1.652,
|
| 3514 |
+
"step": 2400
|
| 3515 |
}
|
| 3516 |
],
|
| 3517 |
"logging_steps": 5,
|