Training in progress, epoch 1

ipu_config.json

@@ -7,7 +7,7 @@
  "execute_encoder_on_cpu_for_generation": false,
  "gradient_accumulation_steps": 32,
  "inference_device_iterations": 4,
- "inference_replication_factor": 1,
+ "inference_replication_factor": 4,
  "ipus_per_replica": 4,
  "layers_per_ipu": [
  3,
@@ -19,11 +19,11 @@
  0.3
  ],
  "optimizer_state_offchip": true,
- "optimum_version": "1.5.1",
+ "optimum_version": "1.5.2",
  "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:a257eeeef9af34da18dbd333cd72aa7f5375b4e43d60c571339096b82e6114d5
-size 171676081
+oid sha256:f61a0b924ad236d4b0756c7caea2890e4b34ae5f309f654ca714691cfb4d9d57
+size 171677805

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f18137acf6e76f22de30f385045a33c51c7e4a1bccce77c50db07e1baaab2313
-size 2799
+oid sha256:8dd38409ca0f12cf3b7b3710b5954ab48e787c3aa818e5d22685fa6d847cf0fc
+size 2811