commit from lixc

README.md

@@ -0,0 +1,16 @@
+
+
+```python
+from transformers import PreTrainedTokenizerFast, BertForMaskedLM
+tokenizer = PreTrainedTokenizerFast.from_pretrained("lixiangchun/transcriptome_iseeek_13millioncells_128tokens")
+iseeek = BertForMaskedLM.from_pretrained("lixiangchun/transcriptome_iseeek_13millioncells_128tokens")
+
+a = ["B2M MTRNR2L8 UBC FOS TMSB4X UBB FTH1 IFITM1 TPT1 FTL DUSP1", "KRT14 MTRNR2L8 KRT6A B2M GAPDH S100A8 S100A9 KRT5"]
+batch = tokenizer(a, max_length=128, truncation=True, padding=True, return_tensors="pt")
+out = iseeek.bert(**batch)
+
+# [CLS] representation
+feature = last_hidden_state[:,0,:]
+
+```
+

config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "/mnt/ssd2/ai/gene_expression/bert/bert_gene_mlm",
+  "architectures": [
+    "BertForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 576,
+  "initializer_range": 0.02,
+  "intermediate_size": 1536,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 384,
+  "model_type": "bert",
+  "num_attention_heads": 8,
+  "num_hidden_layers": 8,
+  "pad_token_id": 3,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.10.0",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 20706
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc8f82423612e372ddcd26ead16df7b2543749e7569b3d455c0e4793fe7ac548
+size 149381452