Update README.md

README.md

@@ -13,15 +13,15 @@ language:
 - km
 - ta
 ---
-# SEA-LION-7B-Instruct-GPTQ
+# SEA-LION-v1-7B-IT-GPTQ
 
 SEA-LION is a collection of Large Language Models (LLMs) which has been pretrained and instruct-tuned for the Southeast Asia (SEA) region.
 The sizes of the models range from 3 billion to 7 billion parameters.
 
-SEA-LION-7B-Instruct is a multilingual model which has been fine-tuned with **thousands of English and Indonesian instruction-completion pairs** alongside a smaller pool of instruction-completion pairs from other ASEAN languages. 
+SEA-LION-v1-7B-IT is a multilingual model which has been fine-tuned with **thousands of English and Indonesian instruction-completion pairs** alongside a smaller pool of instruction-completion pairs from other ASEAN languages. 
 These instructions have been carefully curated and rewritten to ensure the model was trained on truly open, commercially permissive and high quality datasets.
 
-SEA-LION-7B-Instruct-GPTQ is the quantized version of the SEA-LION-7B-Instruct model using a [modified version](https://github.com/caviato/AutoGPTQ) of the [AutoGPTQ](https://github.com/AutoGPTQ/AutoGPTQ) library with Wikipedia texts.
+SEA-LION-v1-7B-IT-GPTQ is the quantized version of SEA-LION-v1-7B-IT using a [modified version](https://github.com/caviato/AutoGPTQ) of the [AutoGPTQ](https://github.com/AutoGPTQ/AutoGPTQ) library with Wikipedia texts.
 
 SEA-LION stands for _Southeast Asian Languages In One Network_.
 
@@ -33,20 +33,20 @@ SEA-LION stands for _Southeast Asian Languages In One Network_.
 
 ## Model Details
 ### Base model 
-SEA-LION-7B-Instruct-GPTQ is quantized from [SEA-LION-7B-Instruct](https://huggingface.co/aisingapore/sea-lion-7b-instruct).
+SEA-LION-v1-7B-IT-GPTQ is quantized from [SEA-LION-v1-7B-IT](https://huggingface.co/aisingapore/SEA-LION-v1-7B-IT).
 
 ### Benchmark Performance
 
 
 | Model                                            | ARC   | HellaSwag | MMLU  | TruthfulQA | Average |
 |--------------------------------------------------|:-----:|:---------:|:-----:|:----------:|:-------:|
-| SEA-LION 7B Instruct (FP16)                      | 40.78 | 68.20     | 27.12 |      36.29 | 43.10   |
-| SEA-LION 7B Instruct GPTQ (4-Bit, 128 group size)| 39.93 | 67.32     | 27.11 |      36.32 | 42.67   |
+| SEA-LION-v1-7B-IT (FP16)                      | 40.78 | 68.20     | 27.12 |      36.29 | 43.10   |
+| SEA-LION-v1-7B-IT-GPTQ (4-Bit, 128 group size)| 39.93 | 67.32     | 27.11 |      36.32 | 42.67   |
 
 ### Usage
 For the full installation, training and inference guide, please refer to the [Github](https://github.com/caviato/sealion-gptq).
 
-In order for SEA-LION-7B-Instruct-GPTQ to work, please install the [modified version of the AutoGPTQ](https://github.com/caviato/AutoGPTQ) library. Installation information can be found [here](https://github.com/caviato/AutoGPTQ#install-from-source).
+In order for SEA-LION-v1-7B-IT-GPTQ to work, please install the [modified version of the AutoGPTQ](https://github.com/caviato/AutoGPTQ) library. Installation information can be found [here](https://github.com/caviato/AutoGPTQ#install-from-source).
 
 SEA-LION can be run using the 🤗 Transformers library 
 ```python
@@ -57,7 +57,7 @@ from auto_gptq import AutoGPTQForCausalLM, BaseQuantizeConfig
 import torch
 
 tokenizer = AutoTokenizer.from_pretrained(
-        "aisingapore/sea-lion-7b-instruct-gptq",
+        "aisingapore/SEA-LION-v1-7B-IT-GBTQ",
         trust_remote_code=True
         )
 
@@ -67,7 +67,7 @@ quantize_config = BaseQuantizeConfig(
         )
 
 model = AutoGPTQForCausalLM.from_quantized( # will be loaded to GPU
-        "aisingapore/sea-lion-7b-instruct-gptq",
+        "aisingapore/SEA-LION-v1-7B-IT-GBTQ",
         device = "cuda:0",
         quantize_config = quantize_config,
         torch_dtype=torch.float16,
@@ -127,10 +127,10 @@ The previous release of the commercially non-permissive SEA-LION-Instruct-Resear
 
 ## Technical Specifications
 ### Fine-Tuning Details
-The SEA-LION-7B-Instruct was fine-tuned using 8x A100-40GB using parameter efficient fine tuning in the form of LoRA.
+SEA-LION-v1-7B-IT was fine-tuned using 8x A100-40GB using parameter efficient fine tuning in the form of LoRA.
 
 ## Data
-SEA-LION-7B-Instruct was trained on a wide range of instructions that were manually and stringently verified by our team. A large portion of the effort was dedicated to ensuring that each instruction-completion pair that the model sees is of a high quality and any errors were corrected and rewritten by native speakers or else dropped from our mix.
+SEA-LION-v1-7B-IT was trained on a wide range of instructions that were manually and stringently verified by our team. A large portion of the effort was dedicated to ensuring that each instruction-completion pair that the model sees is of a high quality and any errors were corrected and rewritten by native speakers or else dropped from our mix.
 
 In addition, special care was taken to ensure that the datasets used had commercially permissive licenses through verification with the original data source.