Upload folder using huggingface_hub

README.md

@@ -0,0 +1,8 @@
+---
+library_name: transformers
+pipeline_tag: text-to-sql
+---
+
+# Text-to-SQL Model
+
+This is a fine-tuned model for converting natural language questions into SQL queries.

chat_template.jinja

@@ -0,0 +1,4 @@
+{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in messages %}{% if message['role'] == 'user' and loop.first %}{{ '### Câu hỏi: ' + message['content'].strip() }}{% elif message['role'] == 'user' %}{{ '
+### Câu hỏi: ' + message['content'].strip() }}{% elif message['role'] == 'assistant' %}{{ '
+### Trả lời: ' + message['content'] + eos_token }}{% endif %}{% if loop.last %}{% if message['role'] == 'user' and add_generation_prompt %}{{ '
+### Trả lời:' }}{% endif %}{% endif %}{% endfor %}

config.json

@@ -0,0 +1,55 @@
+{
+  "architectures": [
+    "MptForCausalLM"
+  ],
+  "attn_config": {
+    "alibi": true,
+    "alibi_bias_max": 8,
+    "attn_impl": "torch",
+    "attn_pdrop": 0.0,
+    "attn_type": "multihead_attention",
+    "attn_uses_sequence_id": false,
+    "clip_qkv": null,
+    "dtype": "float16",
+    "model_type": "",
+    "prefix_lm": false,
+    "qk_ln": false,
+    "softmax_scale": null
+  },
+  "auto_map": {
+    "AutoConfig": "configuration_mpt.MPTConfig",
+    "AutoModelForCausalLM": "modeling_mpt.MPTForCausalLM"
+  },
+  "d_model": 3072,
+  "dtype": "float16",
+  "emb_pdrop": 0.0,
+  "embedding_fraction": 1.0,
+  "expansion_ratio": 4,
+  "init_config": {
+    "emb_init_std": null,
+    "emb_init_uniform_lim": null,
+    "fan_mode": "fan_in",
+    "init_div_is_residual": true,
+    "init_gain": 0.0,
+    "init_nonlinearity": "relu",
+    "init_std": null,
+    "name": "kaiming_normal_",
+    "verbose": 0
+  },
+  "init_device": "cpu",
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "learned_pos_emb": true,
+  "logit_scale": null,
+  "max_seq_len": 8192,
+  "model_type": "mpt",
+  "n_heads": 24,
+  "n_layers": 32,
+  "no_bias": false,
+  "norm_type": "low_precision_layernorm",
+  "resid_pdrop": 0.0,
+  "transformers_version": "4.56.0",
+  "use_cache": false,
+  "verbose": 0,
+  "vocab_size": 20480
+}

generation_config.json

@@ -0,0 +1,5 @@
+{
+  "_from_model_config": true,
+  "transformers_version": "4.56.0",
+  "use_cache": false
+}

model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ddbd754a64a683fcb5c519f46a1b3f60e9069613b6a1506846c8b9394a487f7
+size 4957952664

model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fca83f55e19e638f67a10edca6bed103842cbd4b9859e521a479612f9c6481be
+size 2416055376

model.safetensors.index.json

@@ -0,0 +1,202 @@
+{
+  "metadata": {
+    "total_parameters": 3686992896,
+    "total_size": 7373985792
+  },
+  "weight_map": {
+    "transformer.blocks.0.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.0.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.0.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.0.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.0.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.0.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.1.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.1.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.1.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.1.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.1.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.1.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.10.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.10.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.10.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.10.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.10.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.10.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.11.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.11.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.11.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.11.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.11.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.11.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.12.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.12.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.12.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.12.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.12.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.12.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.13.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.13.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.13.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.13.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.13.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.13.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.14.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.14.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.14.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.14.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.14.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.14.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.15.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.15.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.15.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.15.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.15.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.15.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.16.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.16.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.16.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.16.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.16.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.16.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.17.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.17.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.17.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.17.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.17.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.17.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.18.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.18.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.18.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.18.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.18.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.18.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.19.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.19.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.19.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.19.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.19.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.19.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.2.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.2.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.2.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.2.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.2.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.2.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.20.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.20.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.20.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.20.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.20.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.20.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.21.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.21.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.21.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.21.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.21.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.21.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.22.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.22.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.22.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.22.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.22.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.22.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.23.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.23.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.23.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.23.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.23.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.23.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.24.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.24.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.24.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.24.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.24.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.24.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.25.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.25.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.25.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.25.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.25.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.25.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.26.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.26.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.26.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.26.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.26.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.26.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.27.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.27.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.27.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.27.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.27.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.27.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.28.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.28.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.28.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.28.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.28.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.28.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.29.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.29.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.29.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.29.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.29.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.29.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.3.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.3.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.3.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.3.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.3.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.3.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.30.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.30.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.30.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.30.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.30.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.30.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.31.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.31.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.31.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.31.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.31.norm_1.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.31.norm_2.weight": "model-00002-of-00002.safetensors",
+    "transformer.blocks.4.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.4.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.4.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.4.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.4.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.4.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.5.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.5.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.5.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.5.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.5.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.5.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.6.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.6.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.6.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.6.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.6.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.6.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.7.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.7.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.7.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.7.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.7.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.7.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.8.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.8.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.8.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.8.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.8.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.8.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.9.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.9.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.9.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.9.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.9.norm_1.weight": "model-00001-of-00002.safetensors",
+    "transformer.blocks.9.norm_2.weight": "model-00001-of-00002.safetensors",
+    "transformer.norm_f.weight": "model-00002-of-00002.safetensors",
+    "transformer.wte.weight": "model-00001-of-00002.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,48 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "merges_file": null,
+  "model_max_length": 8192,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "tokenizer_class": "BloomTokenizer",
+  "unk_token": "<unk>",
+  "vocab_file": null
+}