commit files to HF hub

README.md

@@ -0,0 +1,25 @@
+---
+language:
+- en
+tags:
+- openvino
+---
+
+# echarlaix/bert-base-uncased-sst2-acc91.1-d37-hybrid
+
+This is the [echarlaix/bert-base-uncased-sst2-acc91.1-d37-hybrid](https://huggingface.co/echarlaix/bert-base-uncased-sst2-acc91.1-d37-hybrid) model converted to [OpenVINO](https://openvino.ai), for accellerated inference.
+
+An example of how to do inference on this model:
+```python
+from optimum.intel.openvino import OVModelForSequenceClassification
+from transformers import AutoTokenizer, pipeline
+
+# model_id should be set to either a local directory or a model available on the HuggingFace hub.
+model_id = "helenai/echarlaix-bert-base-uncased-sst2-acc91.1-d37-hybrid-ov"
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+model = OVModelForSequenceClassification.from_pretrained(model_id)
+pipe = pipeline("text-classification", model=model, tokenizer=tokenizer)
+result = pipe("I like you. I love you")
+print(result)
+```
+

config.json

@@ -0,0 +1,141 @@
+{
+  "_name_or_path": "echarlaix/bert-base-uncased-sst2-acc91.1-d37-hybrid",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "finetuning_task": "sst2",
+  "gradient_checkpointing": false,
+  "heads_count": 56,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "pruned_heads": {
+    "0": [
+      0,
+      4,
+      6,
+      7,
+      8,
+      9,
+      11
+    ],
+    "1": [
+      0,
+      2,
+      3,
+      5,
+      6,
+      7,
+      8,
+      9
+    ],
+    "2": [
+      2,
+      3,
+      4,
+      5,
+      6,
+      7,
+      8,
+      10,
+      11
+    ],
+    "3": [
+      0,
+      1,
+      2,
+      3,
+      4,
+      6,
+      7,
+      8,
+      11
+    ],
+    "4": [
+      1,
+      2,
+      3,
+      4,
+      8
+    ],
+    "5": [
+      1,
+      2,
+      5,
+      11
+    ],
+    "6": [
+      1,
+      3,
+      7,
+      9,
+      10,
+      11
+    ],
+    "7": [
+      2,
+      4,
+      6,
+      7,
+      11
+    ],
+    "8": [
+      0,
+      2,
+      3,
+      4,
+      5,
+      6
+    ],
+    "9": [
+      0,
+      1,
+      3,
+      4,
+      6,
+      8,
+      9,
+      11
+    ],
+    "10": [
+      0,
+      1,
+      2,
+      3,
+      5,
+      6,
+      7,
+      8,
+      9,
+      10
+    ],
+    "11": [
+      1,
+      2,
+      3,
+      4,
+      5,
+      6,
+      7,
+      8,
+      9,
+      10,
+      11
+    ]
+  },
+  "transformers_version": "4.27.4",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

inference.py

@@ -0,0 +1,10 @@
+from optimum.intel.openvino import OVModelForSequenceClassification
+from transformers import AutoTokenizer, pipeline
+
+# model_id should be set to either a local directory or a model available on the HuggingFace hub.
+model_id = "helenai/echarlaix-bert-base-uncased-sst2-acc91.1-d37-hybrid-ov"
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+model = OVModelForSequenceClassification.from_pretrained(model_id)
+pipe = pipeline("text-classification", model=model, tokenizer=tokenizer)
+result = pipe("I like you. I love you")
+print(result)

openvino_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3037826b4d308719123e419f27d0c7c025d31e0433c3a273be14bebd3da02fc
+size 368272116

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}