Upload GPTNeoXForCausalLM

README.md

@@ -1,74 +1,201 @@
-# KIT-19: A Comprehensive Korean Instruction Toolkit on 19 Tasks for Fine-Tuning Korean Large Language Models
+---
+library_name: transformers
+tags: []
+---
 
-## Introduction
+# Model Card for Model ID
 
-In the current landscape of language models, achieving high performance in Korean NLP tasks requires specialized instruction datasets tailored to the unique aspects of the Korean language. To address the scarcity of such datasets, we introduce **KIT-19**, a comprehensive Korean Instruction Toolkit that encompasses 19 distinct tasks for fine-tuning Korean Large Language Models (LLMs). Unlike existing datasets that largely rely on translated instructions or outputs from models like ChatGPT, KIT-19 is meticulously crafted to capture the nuances of Korean language and culture, offering a robust foundation for advancing Korean LLMs.
+<!-- Provide a quick summary of what the model is/does. -->
 
-## Overview of KIT-19 Datasets
 
-KIT-19 amalgamates 19 existing open-source datasets, each converted into an instruction format to facilitate instruction tuning for Korean LLMs. Here's a brief overview of the datasets included in KIT-19:
 
-| Task Category                         | Datasets Included                                                                                        |
-|---------------------------------------|----------------------------------------------------------------------------------------------------------|
-| Hate Speech Detection                 | APEACH, UnSmile, HateScore                                                                               |
-| Boolean Question Answering (QA)       | KoBEST\_BoolQ                                                                                            |
-| Natural Language Inference (NLI)      | KoBEST\_COPA, korNLI                                                                                     |
-| Text Generation                       | KoBEST\_HellaSwag, kowiki\_text                                                                          |
-| Semantic Textual Similarity (STS)     | korSTS, pawsx\_paraphr, ParaKQC, KoBEST\_WIC, Style\_KQC, Question\_Pair                                 |
-| Sentiment Analysis (SA)               | NSMC                                                                                                     |
-| Intent Argument Extraction            | sae4k\_sum, petitions\_archive                                                                           |
-| Math                                  | math\_korean                                                                                             |
-| Closed Book QA                        | kowiki\_text (utilized differently for Closed Book QA and Text Generation)                              |
-| Summarization                         | lbox\_summarization                                                                                      |
+## Model Details
 
-_Each dataset is selected and formated to ensure a wide coverage of tasks and scenarios relevant to the Korean language, making KIT-19 an exhaustive resource for developing and fine-tuning Korean LLMs._
+### Model Description
 
-## Fine-Tuned Models
+<!-- Provide a longer summary of what this model is. -->
 
-To demonstrate the effectiveness of KIT-19, we have fine-tuned representative Korean Pretrained LLMs, including Polyglot-Ko-5.8b and Polyglot-Ko-1.3b. The fine-tuned models showcase significant performance improvements across a variety of benchmark datasets:
+This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
 
-- KoBEST\_COPA
-- KoBEST\_BoolQ
-- KoBEST\_HellaSwag
-- KLUE\_STS
-- KoBEST\_SentiNeg
-- KLUE\_YNAT
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
 
-The experimental results affirm that **models trained with KIT-19 significantly outperform existing Korean LLMs**, highlighting the potency and necessity of instruction datasets crafted specifically for the Korean language.
+### Model Sources [optional]
 
-# Benchmark Performance
+<!-- Provide the basic links for the model. -->
 
-Below is the performance comparison of different models on various benchmark datasets. The models trained with KIT-19 (KIT-5.8b and KIT-1.3b) are compared against Polyglot-Ko-1.3b, Polyglot-Ko-5.8b, KoAlpaca-5.8b, and Kullm-polyglot-5.8b-v2.
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
 
-| Benchmark Dataset        | Metric      | Polyglot-ko-1.3b | Polyglot-ko-5.8b | KoAlpaca-5.8B | kullm-polyglot-5.8b-v2 | KIT-5.8b       | KIT-1.3b       |
-|--------------------------|-------------|------------------|------------------|---------------|------------------------|----------------|----------------|
-| KoBEST\_COPA             | ACC         | 72.00%           | 77.60%           | 69.80%        | 76.60%                 | **91.60%**     | 83.80%         |
-|                          | F1 (macro)  | 71.96%           | 77.55%           | 69.77%        | 76.53%                 | **91.59%**     | 83.78%         |
-| KoBEST\_BoolQ            | ACC         | 49.86%           | 53.63%           | 56.34%        | 50.28%                 | **66.24%**     | 50.71%         |
-|                          | F1 (macro)  | 35.52%           | 43.56%           | 50.64%        | 33.71%                 | **66.14%**     | 34.78%         |
-| KoBEST\_HellaSwag        | ACC         | 40.60%           | 48.80%           | 38.20%        | 44.40%                 | **97.60%**     | 81.60%         |
-|                          | ACC\_Norm   | 53.00%           | 59.80%           | 46.20%        | 55.20%                 | **98.20%**     | 89.80%         |
-|                          | F1 (macro)  | 40.13%           | 48.53%           | 38.15%        | 44.25%                 | **97.61%**     | 81.49%         |
-| KLUE\_STS                | ACC         | 42.39%           | 45.28%           | 51.83%        | 42.39%                 | **65.51%**     | 42.20%         |
-|                          | F1          | 59.54%           | 60.34%           | 33.86%        | 59.54%                 | **69.71%**     | 56.52%         |
-| KoBEST\_SentiNeg         | ACC         | 69.27%           | 50.38%           | 38.79%        | 50.38%                 | 71.54%         | **80.86%**     |
-|                          | F1          | 68.19%           | 33.95%           | 38.48%        | 33.50%                 | 68.98%         | **80.86%**     |
-| KLUE\_YNAT               | F1          | 33.24%           | 33.62%           | 20.91%        | 32.20%                 | 28.15%         | **38.34%**     |
+## Uses
 
-*Bold* results indicate the best performance in each category.
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
 
-## Conclusion and Future Work
+### Direct Use
 
-KIT-19 stands as a pivotal development in the Korean NLP landscape, addressing the critical need for comprehensive instruction datasets that encapsulate the linguistic and cultural intricacies of the Korean language. With KIT-19, we aim to push the boundaries of what's possible with Korean LLMs, laying a solid foundation for future advancements in the field.
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
 
-We are committed to continuously expanding KIT-19 to cover more domains and further enhance the generability of Korean LLMs. Our hope is that KIT-19 not only serves as a valuable resource for NLP practitioners but also inspires further research and development within the Korean NLP community.
+[More Information Needed]
 
-_The KIT-19 dataset and the fine-tuned models are publicly available for research and development purposes, fueling advancements in Korean language modeling and applications._
+### Downstream Use [optional]
 
----
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure 
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
 
-For more information, access to the datasets, and models, please visit our [GitHub repository](https://github.com/qwer4107/kit-19).
+[More Information Needed]
 
-**Contributors:** Dongjun Jang, Sungjoo Byun, Hyemi Jo, Hyopil Shin from the Department of Linguistics, Seoul National University
 
-_This work is supported by the linguistic insights and technological advances in NLP and aims to contribute to the broader academic and practical applications of language models._

config.json

@@ -0,0 +1,31 @@
+{
+  "_name_or_path": "Junmai/KIT-5.8b",
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "attention_bias": true,
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "classifier_dropout": 0.1,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 16384,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_steps": "global_step320000",
+  "rope_scaling": null,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 0.25,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.37.0",
+  "use_cache": true,
+  "use_parallel_residual": true,
+  "vocab_size": 30080
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 2,
+  "transformers_version": "4.37.0"
+}

model-00001-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5427e8cba6e02fc14863ec758b8e74edf8c0fdf305311c9c162ff3643cbe609
+size 4789002192

model-00002-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:974f157a187eefb59ffb0568013b5fab47f03011bb5218e73e76539c86a1d154
+size 4833124664

model-00003-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd5b18ee4f9114445fa046a73b603c014fecc655cf58acbbff7127bf5b2d6ae2
+size 4833124728

model-00004-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9043040f65c50124c8cf295bede1fe9929df04a8b3d4d781e8e3e2d0e9fafc88
+size 4833124728

model-00005-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f72b054e44520eba831e05dc6374fa97353ea5b32754e280d944db2d616f2b3b
+size 4251900264

model.safetensors.index.json

@@ -0,0 +1,347 @@
+{
+  "metadata": {
+    "total_size": 23540236288
+  },
+  "weight_map": {
+    "embed_out.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.embed_in.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.final_layer_norm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.final_layer_norm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.0.attention.dense.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.attention.dense.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.attention.query_key_value.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.attention.query_key_value.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.input_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.mlp.dense_4h_to_h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.mlp.dense_4h_to_h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.mlp.dense_h_to_4h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.mlp.dense_h_to_4h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.post_attention_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.0.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.attention.dense.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.attention.dense.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.attention.query_key_value.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.attention.query_key_value.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.input_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.mlp.dense_4h_to_h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.mlp.dense_4h_to_h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.mlp.dense_h_to_4h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.mlp.dense_h_to_4h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.post_attention_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.1.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.10.attention.dense.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.attention.dense.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.attention.query_key_value.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.attention.query_key_value.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.input_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.mlp.dense_4h_to_h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.mlp.dense_4h_to_h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.mlp.dense_h_to_4h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.mlp.dense_h_to_4h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.post_attention_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.10.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.attention.dense.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.attention.dense.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.attention.query_key_value.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.attention.query_key_value.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.input_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.mlp.dense_4h_to_h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.11.mlp.dense_4h_to_h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.11.mlp.dense_h_to_4h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.11.mlp.dense_h_to_4h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.11.post_attention_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.11.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.12.attention.dense.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.attention.dense.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.attention.query_key_value.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.attention.query_key_value.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.input_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.mlp.dense_4h_to_h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.mlp.dense_4h_to_h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.mlp.dense_h_to_4h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.mlp.dense_h_to_4h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.post_attention_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.12.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.attention.dense.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.attention.dense.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.attention.query_key_value.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.attention.query_key_value.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.input_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.mlp.dense_4h_to_h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.mlp.dense_4h_to_h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.mlp.dense_h_to_4h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.mlp.dense_h_to_4h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.post_attention_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.13.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.attention.dense.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.attention.dense.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.attention.query_key_value.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.attention.query_key_value.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.input_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.mlp.dense_4h_to_h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.mlp.dense_4h_to_h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.mlp.dense_h_to_4h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.mlp.dense_h_to_4h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.post_attention_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.14.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.attention.dense.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.attention.dense.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.attention.query_key_value.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.attention.query_key_value.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.input_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.mlp.dense_4h_to_h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.mlp.dense_4h_to_h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.mlp.dense_h_to_4h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.mlp.dense_h_to_4h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.post_attention_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.15.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.attention.dense.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.attention.dense.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.attention.query_key_value.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.attention.query_key_value.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.input_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.mlp.dense_4h_to_h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.mlp.dense_4h_to_h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.mlp.dense_h_to_4h.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.mlp.dense_h_to_4h.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.post_attention_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.16.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.attention.dense.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.attention.dense.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.attention.query_key_value.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.attention.query_key_value.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.input_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.input_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.mlp.dense_4h_to_h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.17.mlp.dense_4h_to_h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.17.mlp.dense_h_to_4h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.17.mlp.dense_h_to_4h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.17.post_attention_layernorm.bias": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.17.post_attention_layernorm.weight": "model-00003-of-00005.safetensors",
+    "gpt_neox.layers.18.attention.dense.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.attention.dense.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.attention.query_key_value.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.attention.query_key_value.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.input_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.mlp.dense_4h_to_h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.mlp.dense_4h_to_h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.mlp.dense_h_to_4h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.mlp.dense_h_to_4h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.post_attention_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.18.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.attention.dense.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.attention.dense.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.attention.query_key_value.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.attention.query_key_value.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.input_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.mlp.dense_4h_to_h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.mlp.dense_4h_to_h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.mlp.dense_h_to_4h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.mlp.dense_h_to_4h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.post_attention_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.19.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.2.attention.dense.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.attention.dense.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.attention.query_key_value.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.attention.query_key_value.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.input_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.mlp.dense_4h_to_h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.mlp.dense_4h_to_h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.mlp.dense_h_to_4h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.mlp.dense_h_to_4h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.post_attention_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.2.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.20.attention.dense.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.attention.dense.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.attention.query_key_value.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.attention.query_key_value.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.input_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.mlp.dense_4h_to_h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.mlp.dense_4h_to_h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.mlp.dense_h_to_4h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.mlp.dense_h_to_4h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.post_attention_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.20.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.attention.dense.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.attention.dense.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.attention.query_key_value.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.attention.query_key_value.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.input_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.mlp.dense_4h_to_h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.mlp.dense_4h_to_h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.mlp.dense_h_to_4h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.mlp.dense_h_to_4h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.post_attention_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.21.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.attention.dense.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.attention.dense.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.attention.query_key_value.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.attention.query_key_value.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.input_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.mlp.dense_4h_to_h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.mlp.dense_4h_to_h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.mlp.dense_h_to_4h.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.mlp.dense_h_to_4h.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.post_attention_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.22.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.attention.dense.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.attention.dense.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.attention.query_key_value.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.attention.query_key_value.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.input_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.input_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.mlp.dense_4h_to_h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.23.mlp.dense_4h_to_h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.23.mlp.dense_h_to_4h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.23.mlp.dense_h_to_4h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.23.post_attention_layernorm.bias": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.23.post_attention_layernorm.weight": "model-00004-of-00005.safetensors",
+    "gpt_neox.layers.24.attention.dense.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.attention.dense.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.attention.query_key_value.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.attention.query_key_value.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.input_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.mlp.dense_4h_to_h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.mlp.dense_4h_to_h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.mlp.dense_h_to_4h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.mlp.dense_h_to_4h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.post_attention_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.24.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.attention.dense.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.attention.dense.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.attention.query_key_value.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.attention.query_key_value.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.input_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.mlp.dense_4h_to_h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.mlp.dense_4h_to_h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.mlp.dense_h_to_4h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.mlp.dense_h_to_4h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.post_attention_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.25.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.attention.dense.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.attention.dense.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.attention.query_key_value.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.attention.query_key_value.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.input_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.mlp.dense_4h_to_h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.mlp.dense_4h_to_h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.mlp.dense_h_to_4h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.mlp.dense_h_to_4h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.post_attention_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.26.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.attention.dense.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.attention.dense.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.attention.query_key_value.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.attention.query_key_value.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.input_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.input_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.mlp.dense_4h_to_h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.mlp.dense_4h_to_h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.mlp.dense_h_to_4h.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.mlp.dense_h_to_4h.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.post_attention_layernorm.bias": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.27.post_attention_layernorm.weight": "model-00005-of-00005.safetensors",
+    "gpt_neox.layers.3.attention.dense.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.attention.dense.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.attention.query_key_value.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.attention.query_key_value.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.input_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.mlp.dense_4h_to_h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.mlp.dense_4h_to_h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.mlp.dense_h_to_4h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.mlp.dense_h_to_4h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.post_attention_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.3.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.attention.dense.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.attention.dense.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.attention.query_key_value.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.attention.query_key_value.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.input_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.mlp.dense_4h_to_h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.mlp.dense_4h_to_h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.mlp.dense_h_to_4h.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.mlp.dense_h_to_4h.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.post_attention_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.4.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.attention.dense.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.attention.dense.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.attention.query_key_value.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.attention.query_key_value.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.input_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.input_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.mlp.dense_4h_to_h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.5.mlp.dense_4h_to_h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.5.mlp.dense_h_to_4h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.5.mlp.dense_h_to_4h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.5.post_attention_layernorm.bias": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.5.post_attention_layernorm.weight": "model-00001-of-00005.safetensors",
+    "gpt_neox.layers.6.attention.dense.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.attention.dense.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.attention.query_key_value.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.attention.query_key_value.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.input_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.mlp.dense_4h_to_h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.mlp.dense_4h_to_h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.mlp.dense_h_to_4h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.mlp.dense_h_to_4h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.post_attention_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.6.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.attention.dense.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.attention.dense.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.attention.query_key_value.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.attention.query_key_value.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.input_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.mlp.dense_4h_to_h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.mlp.dense_4h_to_h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.mlp.dense_h_to_4h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.mlp.dense_h_to_4h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.post_attention_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.7.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.attention.dense.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.attention.dense.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.attention.query_key_value.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.attention.query_key_value.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.input_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.mlp.dense_4h_to_h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.mlp.dense_4h_to_h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.mlp.dense_h_to_4h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.mlp.dense_h_to_4h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.post_attention_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.8.post_attention_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.attention.dense.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.attention.dense.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.attention.query_key_value.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.attention.query_key_value.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.input_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.input_layernorm.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.mlp.dense_4h_to_h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.mlp.dense_4h_to_h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.mlp.dense_h_to_4h.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.mlp.dense_h_to_4h.weight": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.post_attention_layernorm.bias": "model-00002-of-00005.safetensors",
+    "gpt_neox.layers.9.post_attention_layernorm.weight": "model-00002-of-00005.safetensors"
+  }
+}