Training in progress, step 500

config.json

@@ -24,6 +24,7 @@
   "num_expert_modules": 3,
   "num_experts": 8,
   "num_hidden_layers": 16,
+  "output_hidden_states": true,
   "pad_token_id": 0,
   "router_jitter_noise": 0.01,
   "top_k": 2,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a08a3eb99ffc4af8ded28b27670cdcf6db889a2efe9f0a096387818a4d82102f
-size 943411142
+oid sha256:503df7e488f14fdbf767b050f9c5b78af673b804625fe836d9bd1d7886e8511f
+size 1059459406

runs/Sep14_22-18-51_nid007230/events.out.tfevents.1757881139.nid007230.6286.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:49cc910fd5f5175c979f04810b0db31545c6f6953b68d22c1ba956e96851848b
-size 396385
+oid sha256:58ddf885b338ba0bbe6a3a2a0efa6304b53dba9bceaafd421985b905e86a8ef3
+size 403050

runs/Sep15_09-51-02_nid007560/events.out.tfevents.1757922679.nid007560.261976.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d5acb1e9be0e3c476f98d1fd3a9fb0a30171cf3a46e9e89e574619dc9eba2c9
+size 8774

runs/Sep15_11-38-51_nid007040/events.out.tfevents.1757929138.nid007040.116713.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3172d87c371c7b9aac16b1391d82dad625281b9f7e3972a6e73f81a9f3a2f69e
+size 15737

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9540043b69e398857104c563a1ec22c9bcc89baa9d452eba21b4d93a7a7d15b2
+oid sha256:c1f37caca9737ccfc98bf318f083d1e9f58e8000a73e07781b9abf5ca65a160a
 size 5432