First

.gitattributes

@@ -0,0 +1 @@
+*.safetensors filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,169 @@
+---
+language:
+- en
+license: other
+library_name: robotstar
+pipeline_tag: robotics
+tags:
+- sign-language-generation
+- text-to-motion
+- robotics
+- motion-generation
+datasets:
+- how2sign
+---
+
+<h1 align="center"><strong>RobotSTAR: Next-Scale Autoregressive Sign Language Motion Translation for Humanoid Robots</strong></h1>
+  <p align="center">
+              <a href="https://www.yujiazeng.com/">Yujia Zeng<sup>*</sup></a>
+              <a href="https://pengchensheng.com/">Chensheng Peng<sup>*</sup></a>
+              <a href="https://thomaschen98.github.io/">Yuxin Chen<sup>*</sup></a>
+              <a href="https://www.yujiazeng.com/">xxx1<sup></sup></a>
+              <a href="https://www.yujiazeng.com/">xxx2<sup></sup></a>
+              <a href="https://me.berkeley.edu/people/masayoshi-tomizuka/">Masayoshi Tomizuka<sup></sup></a>    <br>
+    <sup></sup>MSC LAB, UC Berkeley <br> <sup>*</sup>Equal contribution
+</p>
+
+<p align="center">
+  <img src="https://img.shields.io/badge/arXiv-Coming_Soon-blue">
+  <img src="https://img.shields.io/badge/Project-Coming_Soon-blue">
+  <img src="https://img.shields.io/badge/🤗_HuggingFace-Coming_Soon-orange">
+</p>
+
+<p align="center">
+  <em>
+    RobotSTAR translates English text or audio into continuous American Sign Language (ASL) motion. <br>
+    This release contains the text-to-sign-motion component.
+  </em>
+</p>
+
+
+## Installation
+
+
+```bash
+git clone https://github.com/zyjOrz/RobotSTAR.git
+cd RobotSTAR
+conda create -n robotstar python=3.10 -y
+conda activate robotstar
+pip install -e .
+```
+
+
+## Inference
+
+Predicted-duration inference:
+
+```bash
+python -m robotstar.infer \
+  --model Ivystream/RobotSTAR \
+  --text "A person explains the plan." \
+  --length-mode predicted \
+  --output-dir outputs/predicted
+```
+
+For a stable demonstration duration, specify seconds explicitly:
+
+```bash
+python -m robotstar.infer \
+  --model Ivystream/RobotSTAR \
+  --text "A person explains the plan." \
+  --length-mode seconds \
+  --length-value 6.0 \
+  --output-dir outputs/six_seconds
+```
+
+Other supported modes are `tokens` and `frames`. 
+
+
+## Data preparation
+
+
+```bash
+python -m robotstar.prepare_data \
+  --train-manifest data/train_raw.jsonl \
+  --val-manifest data/val_raw.jsonl \
+  --test-manifest data/test_raw.jsonl \
+  --output data/prepared
+```
+
+## Training
+
+### 1. Train the FSQ tokenizer
+
+```bash
+torchrun --standalone --nproc_per_node=8 -m robotstar.train_tokenizer \
+  --config configs/tokenizer.yaml \
+  --prepared-root data/prepared \
+  --output experiments/robotstar_tokenizer
+```
+
+### 2. Export motion tokens
+
+```bash
+torchrun --standalone --nproc_per_node=8 -m robotstar.export_tokens \
+  --model experiments/robotstar_tokenizer/best.pt \
+  --prepared-root data/prepared \
+  --output data/tokens
+```
+
+### 3. Build the optional retrieval memory
+
+The isolated-word source used in our experiments is [`akasheroor/American-Sign-Language-Dataset`](https://huggingface.co/datasets/akasheroor/American-Sign-Language-Dataset).
+
+```bash
+python -m robotstar.retrieval build \
+  --word-token-jsonl data/word_tokens/train_source_tokens.jsonl \
+  --output data/retrieval/word2code.json
+```
+
+
+### 4. Build coarse-to-fine caches
+
+```bash
+python -m robotstar.build_cache \
+  --token-root data/tokens \
+  --tokenizer-model experiments/robotstar_tokenizer/best.pt \
+  --base-model google/mt5-large \
+  --retrieval data/retrieval/word2code.json \
+  --output data/cache
+```
+
+### 5. Train RobotSTAR
+
+```bash
+torchrun --standalone --nproc_per_node=8 -m robotstar.train_generator \
+  --config configs/robotstar_mt5_large.yaml \
+  --cache-root data/cache \
+  --output experiments/robotstar_mt5_large
+```
+
+
+## Evaluation
+
+```bash
+python -m robotstar.evaluate \
+  --predictions outputs/predictions \
+  --manifest data/prepared/test.jsonl
+```
+
+
+## 🔗 Citation
+
+If you find our work useful, please consider citing:
+
+```bibtex
+@article{robotstar,
+  title   = {RobotSTAR: ...},
+  author  = {...},
+  journal = {...},
+  year    = {2026}
+}
+```
+
+## 🙏 Acknowledgements
+
+- [SOKE](https://github.com/2000ZRL/SOKE): Our work follows SOKE in adopting its motion representation and sign retrieval formulation.
+
+- [How2Sign Dataset](https://how2sign.github.io/): Our work uses the How2Sign dataset for training and evaluation. 
+

generator/checkpoint_metadata.json

@@ -0,0 +1,8 @@
+{
+  "version": "robotstar_fsq_c2f_mt5_large_v1",
+  "source_checkpoint_sha256": "429b4b3948270d9216d6f3c90e3eb2bf60f9b1ba8fdbd9b5dd2233801acd963d",
+  "step": 894,
+  "epoch": 6,
+  "optimizer_removed": true,
+  "scaler_removed": true
+}

generator/config.json

@@ -0,0 +1,32 @@
+{
+  "body_codes": 96,
+  "hand_codes": 192,
+  "alpha_hand": 0.4,
+  "max_source_length": 512,
+  "max_motion_tokens": 100,
+  "use_compact_map": true,
+  "attn_implementation": "eager",
+  "stage_divisors": [
+    8,
+    4,
+    2,
+    1
+  ],
+  "label_smoothing": 0.0,
+  "stage_loss_weights": [
+    1.25,
+    1.1,
+    1.0,
+    1.25
+  ],
+  "hand_loss_weight": 1.25,
+  "length_loss_weight": 0.2,
+  "contrastive_loss_weight": 0.03,
+  "contrastive_temperature": 0.07,
+  "lm_family": "mt5",
+  "model_path": "google/mt5-large",
+  "token_namespace": "robotstar",
+  "scale_decoder_outputs": true,
+  "public_model_name": "RobotSTAR",
+  "format_version": "robotstar_generator_config_v1_1"
+}

generator/model-00001-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f3ade9d1453c35b03f5abca1c3a67d4448a5e4cec670c06709b2557f6510ac2
+size 1657530872

generator/model-00002-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f45f31f453df8259ab727b2ff9e1f4f63ce214898a6327dbf0fea1fbf9dc53a7
+size 1991250056

generator/model-00003-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d39a1bd2d2d0f86982d7a21670abb49c0d51539c4f410422559f8b0a5e42c268
+size 1294870664

generator/model-00004-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3fe0fb77d218b4dcb38e7fe7f8011d28ad0f5605e368b9e85044fd4a6374909
+size 1026379912

generator/model-00005-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e16c106cae12b94ab9a88c42414f75cb7e44f5bb2fe5cc1e9a6fc7376bbebdc3
+size 1034785208

generator/model.safetensors.index.json

@@ -0,0 +1,582 @@
+{
+  "metadata": {
+    "total_size": 7004742096
+  },
+  "weight_map": {
+    "coarse_seed": "model-00001-of-00005.safetensors",
+    "context_projection.0.weight": "model-00001-of-00005.safetensors",
+    "context_projection.1.weight": "model-00001-of-00005.safetensors",
+    "context_projection.2.weight": "model-00001-of-00005.safetensors",
+    "context_projection.3.weight": "model-00001-of-00005.safetensors",
+    "cross_scale_gates": "model-00001-of-00005.safetensors",
+    "length_head.bias": "model-00001-of-00005.safetensors",
+    "length_head.weight": "model-00001-of-00005.safetensors",
+    "length_pool.0.bias": "model-00001-of-00005.safetensors",
+    "length_pool.0.weight": "model-00001-of-00005.safetensors",
+    "length_pool.1.bias": "model-00001-of-00005.safetensors",
+    "length_pool.1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.0.SelfAttention.relative_attention_bias.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.0.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.1.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.10.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.11.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.12.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.13.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.14.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.15.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.16.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.17.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.18.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.19.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.2.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.20.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.21.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.22.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.23.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.3.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.4.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.5.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.6.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.7.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.8.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.0.SelfAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.0.SelfAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.0.SelfAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.0.SelfAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.0.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.1.EncDecAttention.k.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.1.EncDecAttention.o.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.1.EncDecAttention.q.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.1.EncDecAttention.v.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.1.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.2.DenseReluDense.wi_0.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.2.DenseReluDense.wi_1.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.2.DenseReluDense.wo.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.block.9.layer.2.layer_norm.weight": "model-00001-of-00005.safetensors",
+    "main_lm.decoder.embed_tokens.weight": "model-00002-of-00005.safetensors",
+    "main_lm.decoder.final_layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.0.SelfAttention.relative_attention_bias.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.0.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.1.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.10.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.11.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.12.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.13.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.14.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.15.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.16.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.17.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.18.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.19.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.2.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.20.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.21.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.22.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.23.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.1.DenseReluDense.wo.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.3.layer.1.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.0.SelfAttention.k.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.0.SelfAttention.o.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.0.SelfAttention.q.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.0.SelfAttention.v.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.0.layer_norm.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.1.DenseReluDense.wi_0.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.1.DenseReluDense.wi_1.weight": "model-00002-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.1.DenseReluDense.wo.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.4.layer.1.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.0.SelfAttention.k.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.0.SelfAttention.o.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.0.SelfAttention.q.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.0.SelfAttention.v.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.0.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.1.DenseReluDense.wi_0.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.1.DenseReluDense.wi_1.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.1.DenseReluDense.wo.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.5.layer.1.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.0.SelfAttention.k.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.0.SelfAttention.o.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.0.SelfAttention.q.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.0.SelfAttention.v.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.0.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.1.DenseReluDense.wi_0.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.1.DenseReluDense.wi_1.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.1.DenseReluDense.wo.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.6.layer.1.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.0.SelfAttention.k.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.0.SelfAttention.o.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.0.SelfAttention.q.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.0.SelfAttention.v.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.0.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.1.DenseReluDense.wi_0.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.1.DenseReluDense.wi_1.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.1.DenseReluDense.wo.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.7.layer.1.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.0.SelfAttention.k.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.0.SelfAttention.o.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.0.SelfAttention.q.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.0.SelfAttention.v.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.0.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.1.DenseReluDense.wi_0.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.1.DenseReluDense.wi_1.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.1.DenseReluDense.wo.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.8.layer.1.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.0.SelfAttention.k.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.0.SelfAttention.o.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.0.SelfAttention.q.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.0.SelfAttention.v.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.0.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.1.DenseReluDense.wi_0.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.1.DenseReluDense.wi_1.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.1.DenseReluDense.wo.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.block.9.layer.1.layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.embed_tokens.weight": "model-00003-of-00005.safetensors",
+    "main_lm.encoder.final_layer_norm.weight": "model-00003-of-00005.safetensors",
+    "main_lm.lm_head.weight": "model-00004-of-00005.safetensors",
+    "main_lm.shared.weight": "model-00005-of-00005.safetensors",
+    "motion_contrastive.weight": "model-00005-of-00005.safetensors",
+    "stage_embedding.weight": "model-00005-of-00005.safetensors",
+    "text_contrastive.weight": "model-00005-of-00005.safetensors"
+  }
+}

generator/vocab.json

@@ -0,0 +1,28 @@
+{
+  "model_path": "google/mt5-large",
+  "body_codes": 96,
+  "hand_codes": 192,
+  "use_compact_map": true,
+  "vocab_size": 250581,
+  "pad_id": 0,
+  "eos_id": 1,
+  "en_asl_id": 250100,
+  "token_namespace": "robotstar",
+  "token_contract": {
+    "version": "robotstar_fsq_c2f_mt5_single_code_v1",
+    "frame_dim": 133,
+    "body_dim": 43,
+    "left_dim": 45,
+    "right_dim": 45,
+    "body_vocab": 96,
+    "hand_vocab": 192,
+    "stage_divisors": [
+      8,
+      4,
+      2,
+      1
+    ],
+    "temporal_downsample": 4,
+    "token_namespace": "robotstar"
+  }
+}

tokenizer/checkpoint_metadata.json

@@ -0,0 +1,8 @@
+{
+  "version": "robotstar_fsq_tokenizer_v1",
+  "source_checkpoint_sha256": "c3520f377a8fb4e0ad8cdf4472c9817e71397fee1883f66b8da59ae60bd99293",
+  "step": 140,
+  "epoch": 10,
+  "optimizer_removed": true,
+  "scaler_removed": true
+}

tokenizer/config.json

@@ -0,0 +1,22 @@
+{
+  "body_code_num": 96,
+  "hand_code_num": 192,
+  "body_levels": [
+    4,
+    4,
+    6
+  ],
+  "hand_levels": [
+    4,
+    6,
+    8
+  ],
+  "code_dim": 512,
+  "width": 512,
+  "depth": 3,
+  "down_t": 2,
+  "stride_t": 2,
+  "dilation_growth_rate": 3,
+  "public_model_name": "RobotSTAR",
+  "format_version": "robotstar_tokenizer_config_v1"
+}

tokenizer/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbed6aea2173ba27a81695de45e2c00569e292e8520a66658bb262c375ac103f
+size 225244248