Training in progress, step 100000

blimp_results.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76410b857f71ec7f1878911728d5905774d47f099712cf67110325d7a8caad09
-size 80230056
+oid sha256:212130ec3f4bb86bfc2b9749511392097eaefc6d4f50037b134a632f85d17cb1
+size 80216458

config.json

@@ -3,20 +3,20 @@
   "architectures": [
     "GPT2LMHeadModel"
   ],
-  "attn_pdrop": 0.3,
+  "attn_pdrop": 0.1,
   "bos_token_id": 0,
-  "embd_pdrop": 0.3,
+  "embd_pdrop": 0.1,
   "eos_token_id": 0,
   "initializer_range": 0.02,
   "layer_norm_epsilon": 1e-05,
   "model_type": "gpt2",
-  "n_embd": 768,
-  "n_head": 12,
-  "n_inner": 3072,
-  "n_layer": 12,
+  "n_embd": 512,
+  "n_head": 8,
+  "n_inner": 2048,
+  "n_layer": 6,
   "n_positions": 256,
   "reorder_and_upcast_attn": false,
-  "resid_pdrop": 0.3,
+  "resid_pdrop": 0.1,
   "scale_attn_by_inverse_layer_idx": false,
   "scale_attn_weights": true,
   "summary_activation": null,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c3f57720261820330ee95da753a6c15ae9d37bb82e68ac923e61276d685a882
-size 390177408
+oid sha256:670bbe67b4ddafe189a296862499109994dab1c512f310dceaba7af83b281ea0
+size 108961160

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6bf434d08adffb63320a056f43c3d312c362758a72db108fcc5c86bb963bff1
+oid sha256:5dc80136f7f23f477ebdd661d182228c4de21c404425ed1d103d090bb1541237
 size 5368