Training in progress, step 500

config.json

@@ -20,7 +20,7 @@
   "pad_token_id": 1,
   "position_embedding_type": "absolute",
   "torch_dtype": "float16",
-  "transformers_version": "4.24.0",
+  "transformers_version": "4.20.1",
   "type_vocab_size": 1,
   "use_cache": true,
   "vocab_size": 50265

ipu_config.json

@@ -1,8 +1,7 @@
 {
-  "decompose_grad_sum": false,
+  "auto_loss_scaling": false,
   "device_iterations": 1,
   "embedding_serialization_factor": 0,
-  "enable_half_first_order_momentum": true,
   "enable_half_partials": true,
   "executable_cache_dir": "/tmp/exe_cache",
   "execute_encoder_on_cpu_for_generation": false,
@@ -20,14 +19,12 @@
     0.25
   ],
   "optimizer_state_offchip": true,
-  "optimum_version": "1.4.1",
+  "optimum_version": "1.5.0",
   "output_mode": "final",
-  "profile_dir": "",
   "recompute_checkpoint_every_layer": true,
   "replicated_tensor_sharding": false,
   "replication_factor": 1,
   "seed": 42,
   "sharded_execution_for_inference": false,
-  "transformers_version": "4.24.0",
-  "use_popdist": false
+  "transformers_version": "4.20.1"
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:47ea93dbcc63d6c53bc5d3a8a9b6df66a62185083bf6e787b831b1f036039669
+oid sha256:33c1379f33bd67409aedabb3c7779827cdadfa8268a8649b3c74b5daa2f08083
 size 249361645

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:289af17c2cc930989397aff4a5aa4b60661acf488970848eb18b1b14ae06a2d1
-size 2671
+oid sha256:1440d9b094939f46d60953bf6b35a663a09e723d5df1530c1084afb6bc78115c
+size 2735