Upload GLUE result: COLA-0_lr_only

results/glue/COLA-0_lr_only/config.json

@@ -0,0 +1,92 @@
+{
+  "attention_probs_dropout_prob": 0.1,
+  "cola_mlp": true,
+  "config_forward": null,
+  "config_tensor": {
+    "attn": {
+      "fc": {
+        "amp": true,
+        "build_rank_parameters": false,
+        "config_forward": null,
+        "factorization": "tt",
+        "lr_act": "gelu",
+        "per_decomp_rank_ratio_limit": 1.0,
+        "ranks": 384,
+        "set_scale_factors": false,
+        "shape": [
+          1024,
+          1024
+        ],
+        "tensorized": true
+      },
+      "sub_layers": {
+        "amp": true,
+        "build_rank_parameters": false,
+        "config_forward": null,
+        "factorization": "tt",
+        "lr_act": "gelu",
+        "per_decomp_rank_ratio_limit": 1.0,
+        "ranks": 384,
+        "set_scale_factors": false,
+        "shape": [
+          1024,
+          1024
+        ],
+        "tensorized": true
+      }
+    },
+    "embedding": {
+      "tensorized": false
+    },
+    "head": {
+      "tensorized": false
+    },
+    "pff": [
+      {
+        "amp": true,
+        "build_rank_parameters": false,
+        "config_forward": null,
+        "factorization": "tt",
+        "lr_act": "gelu",
+        "per_decomp_rank_ratio_limit": 1.0,
+        "ranks": 512,
+        "set_scale_factors": false,
+        "shape": [
+          1024,
+          2736
+        ],
+        "tensorized": true
+      },
+      {
+        "amp": true,
+        "build_rank_parameters": false,
+        "config_forward": null,
+        "factorization": "tt",
+        "lr_act": "gelu",
+        "per_decomp_rank_ratio_limit": 1.0,
+        "ranks": 512,
+        "set_scale_factors": false,
+        "shape": [
+          2736,
+          1024
+        ],
+        "tensorized": true
+      }
+    ],
+    "pooler": {
+      "tensorized": false
+    }
+  },
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 2736,
+  "max_position_embeddings": 512,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "only_lr_act": true,
+  "output_all_encoded_layers": false,
+  "type_vocab_size": 2,
+  "vocab_size": 30528
+}

results/glue/COLA-0_lr_only/dllogger.json

@@ -0,0 +1,13 @@
+DLLL {"timestamp": "1755868214.299715", "elapsedtime": "3e-06", "datetime": "2025-08-22 13:10:14.299715+00:00", "type": "METADATA", "metric": "e2e_train_time", "metadata": {"unit": "s"}}
+DLLL {"timestamp": "1755868214.299748", "elapsedtime": "3.6e-05", "datetime": "2025-08-22 13:10:14.299748+00:00", "type": "METADATA", "metric": "training_sequences_per_second", "metadata": {"unit": "sequences/s"}}
+DLLL {"timestamp": "1755868214.299759", "elapsedtime": "4.7e-05", "datetime": "2025-08-22 13:10:14.299759+00:00", "type": "METADATA", "metric": "e2e_inference_time", "metadata": {"unit": "s"}}
+DLLL {"timestamp": "1755868214.299767", "elapsedtime": "5.5e-05", "datetime": "2025-08-22 13:10:14.299767+00:00", "type": "METADATA", "metric": "inference_sequences_per_second", "metadata": {"unit": "sequences/s"}}
+DLLL {"timestamp": "1755868214.299773", "elapsedtime": "6.1e-05", "datetime": "2025-08-22 13:10:14.299773+00:00", "type": "METADATA", "metric": "exact_match", "metadata": {"unit": null}}
+DLLL {"timestamp": "1755868214.299781", "elapsedtime": "6.9e-05", "datetime": "2025-08-22 13:10:14.299781+00:00", "type": "METADATA", "metric": "F1", "metadata": {"unit": null}}
+DLLL {"timestamp": "1755868214.29981", "datetime": "2025-08-22 13:10:14.299810+00:00", "elapsedtime": "9.8e-05", "type": "LOG", "step": "PARAMETER", "data": {"Config": ["Namespace(data_dir='/datasets/glue/COLA/', bert_model='bert-large-uncased', task_name='cola', output_dir='tensor-/results/glue/COLA-0_lr_only', init_checkpoint='/results/cola_bert_llama_mlp_large_0.7_only_lr-001/model.pt', max_seq_length=128, do_train=True, do_eval=True, do_predict=False, do_lower_case=True, train_batch_size=128, eval_batch_size=8, learning_rate=2.4e-05, num_train_epochs=3.0, max_steps=-1.0, warmup_proportion=0.1, no_cuda=False, local_rank=-1, seed=0, gradient_accumulation_steps=32, fp16=True, amp=False, loss_scale=0, server_ip='', server_port='', vocab_file='/workspace/TensorBERT-FT/vocab', config_file='/results/cola_bert_llama_mlp_large_0.7_only_lr-001', skip_checkpoint=False, rank_adaptive=False, no_update_factors=False, rank_learning_rate=None, rank_lr_schedule='constant_warmup', disable_progress_bar=False)"]}}
+DLLL {"timestamp": "1755868214.300177", "datetime": "2025-08-22 13:10:14.300177+00:00", "elapsedtime": "0.000465", "type": "LOG", "step": "PARAMETER", "data": {"SEED": 0}}
+DLLL {"timestamp": "1755868220.941711", "datetime": "2025-08-22 13:10:20.941711+00:00", "elapsedtime": "6.641999", "type": "LOG", "step": "PARAMETER", "data": {"num_parameters": 247299330}}
+DLLL {"timestamp": "1755868479.146081", "datetime": "2025-08-22 13:14:39.146081+00:00", "elapsedtime": "264.846369", "type": "LOG", "step": [], "data": {"e2e_train_time": 254.42025417089462}}
+DLLL {"timestamp": "1755868479.146124", "datetime": "2025-08-22 13:14:39.146124+00:00", "elapsedtime": "264.846412", "type": "LOG", "step": [], "data": {"training_sequences_per_second": 100.82923658573513}}
+DLLL {"timestamp": "1755868479.146139", "datetime": "2025-08-22 13:14:39.146139+00:00", "elapsedtime": "264.846427", "type": "LOG", "step": [], "data": {"e2e_inference_time": 2.0460436782836915}}
+DLLL {"timestamp": "1755868479.146153", "datetime": "2025-08-22 13:14:39.146153+00:00", "elapsedtime": "264.846441", "type": "LOG", "step": [], "data": {"inference_sequences_per_second": 512.2080291458426}}

results/glue/COLA-0_lr_only/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:567e733a3c37f22660c4262e8dcc0baff8b265893c50e2c540acf1fda2859ea0
+size 989482894

results/glue/COLA-0_lr_only/results.txt

@@ -0,0 +1 @@
+{"global_step": 198, "train:loss": 0.004139550511560915, "train:latency": 254.42025417089462, "train:num_samples_per_gpu": 25653, "train:num_steps": 2138, "train:throughput": 100.82923658573513, "eval:num_samples_per_gpu": 1043, "eval:num_steps": 131, "infer:latency(ms):50%": 15.344223976135254, "infer:latency(ms):90%": 16.5928955078125, "infer:latency(ms):95%": 16.614240646362305, "infer:latency(ms):99%": 16.685056686401367, "infer:latency(ms):100%": 16.99942398071289, "infer:latency(ms):avg": 15.618654032699935, "infer:latency(ms):std": 0.6911833037440174, "infer:latency(ms):sum": 2046.0436782836914, "infer:throughput(samples/s):avg": 512.2080291458426, "eval:loss": 0.7501575054106736, "mcc": 0.4519141240925685}