Training in progress, step 140000

config.json

@@ -3,9 +3,9 @@
   "architectures": [
     "GPT2LMHeadModel"
   ],
-  "attn_pdrop": 0.3,
+  "attn_pdrop": 0.1,
   "bos_token_id": 3,
-  "embd_pdrop": 0.3,
+  "embd_pdrop": 0.1,
   "eos_token_id": 3,
   "initializer_range": 0.02,
   "layer_norm_epsilon": 1e-05,
@@ -13,10 +13,10 @@
   "n_embd": 128,
   "n_head": 4,
   "n_inner": 512,
-  "n_layer": 3,
+  "n_layer": 2,
   "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 sha256:6aaada6e2c6ec557ad1aadfc3d577dc8f31c19ac223931ea970e592fe7ddf841
-size 2539288
+oid sha256:cb552f4f853292007f968542924e7cc5be6d717d80d068cddd1dfeae54e0a1ba
+size 1745032

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:82ae76f4757a61676b98de487ca62d4fd33f207f408709e8e1ffbe6551050475
-size 5368
+oid sha256:35d7abef20837f2b1d352ffbb2d704a4f9673b753267189e093b1925e12e7d98
+size 5432