Training in progress, step 500

config.json

@@ -1,7 +1,7 @@
 {
   "_name_or_path": "bert-base-cased",
   "architectures": [
-    "PipelinedBertForMaskedLM"
+    "PoptorchPipelinedBertForMaskedLM"
   ],
   "attention_probs_dropout_prob": 0.1,
   "classifier_dropout": null,

ipu_config.json

@@ -7,7 +7,7 @@
   "execute_encoder_on_cpu_for_generation": false,
   "gradient_accumulation_steps": 64,
   "inference_device_iterations": 5,
-  "inference_replication_factor": 1,
+  "inference_replication_factor": 4,
   "ipus_per_replica": 4,
   "layers_per_ipu": [
     0,
@@ -20,8 +20,8 @@
   "optimum_version": "1.5.1",
   "output_mode": "final",
   "recompute_checkpoint_every_layer": true,
-  "replicated_tensor_sharding": false,
-  "replication_factor": 1,
+  "replicated_tensor_sharding": true,
+  "replication_factor": 4,
   "seed": 42,
   "sharded_execution_for_inference": false,
   "transformers_version": "4.20.1"

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1de78a8b8fb98abdcd2dbbf6829cfe16ec3b973772d87bc9706197039ebfc620
+oid sha256:8dda0f0e8c290dcaedb6403fc755cb91b3b0416cb726e5e23d058a6fafa30e73
 size 216750379

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc20b1757df236592af55aae1d45cfd378146a983b8d10058217d3d328147e36
+oid sha256:ae44d914d31acb80a2e57e6b15462f3e308d9464af4f280fad43e367db1f742e
 size 2735