diff --git a/.gitattributes b/.gitattributes
index 80de38108a691c19013832938c08aa6be4d0d4f0..1ed8110b98f923d72d3930d996586a6e538d842a 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -70,3 +70,18 @@ adapters/proofwriter-demonstration-self_supervised/tokenizer.json filter=lfs dif
 adapters/musique-rephrase-self_unsupervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 adapters/musique-demonstration-self_supervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 adapters/musique-feedback-external/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-demonstration-self_supervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/proofwriter-rephrase-external/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/proofwriter-feedback-self_supervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-rephrase-self_unsupervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/proofwriter-rephrase-self_supervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-demonstration-self_unsupervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/proofwriter-feedback-external/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-rephrase-external/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-feedback-self_unsupervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/proofwriter-rephrase-self_unsupervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/proofwriter-feedback-self_unsupervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-demonstration-external/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-feedback-external/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-feedback-self_supervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+adapters/big_math-rephrase-self_supervised/tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/adapters/big_math-demonstration-external/README.md b/adapters/big_math-demonstration-external/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..eb09148de648cee968a2971cee1f2dc513f83603
--- /dev/null
+++ b/adapters/big_math-demonstration-external/README.md
@@ -0,0 +1,207 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **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]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### 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
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/adapters/big_math-demonstration-external/adapter_config.json b/adapters/big_math-demonstration-external/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0ed2f182b878c5aebb8041ea17e56a50d8742fe
--- /dev/null
+++ b/adapters/big_math-demonstration-external/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
+  "target_modules": [
+    "down_proj",
+    "up_proj",
+    "v_proj",
+    "q_proj",
+    "k_proj",
+    "o_proj",
+    "gate_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
diff --git a/adapters/big_math-demonstration-external/adapter_model.safetensors b/adapters/big_math-demonstration-external/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..57103bed3d65e38a9b88cbb09ac45acb1b66bba6
--- /dev/null
+++ b/adapters/big_math-demonstration-external/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0bdbcb63bd9ee8dedb34b30f92e6169266bd38d1f93e325e42b19ffc4419f82
+size 161533192
diff --git a/adapters/big_math-demonstration-external/chat_template.jinja b/adapters/big_math-demonstration-external/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e
--- /dev/null
+++ b/adapters/big_math-demonstration-external/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/adapters/big_math-demonstration-external/resolved_training_config.json b/adapters/big_math-demonstration-external/resolved_training_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c39ef9c763de4bf7b694ccbf555426f76e3ab8d8
--- /dev/null
+++ b/adapters/big_math-demonstration-external/resolved_training_config.json
@@ -0,0 +1,39 @@
+{
+  "condition": "big_math-demonstration-external",
+  "context_manifest_sha256": "0c21c0c76d1ad8dae8ae8fba0336fc709b369e9b71d9a7020f812e1a51dde474",
+  "manifest_sha256": "bc3cad94dbd82970520e3c1c2964799c91b7550a7457f076b46f0aae2ce0ed6a",
+  "model": {
+    "repo": "Qwen/Qwen2.5-7B-Instruct",
+    "revision": "a09a35458c702b33eeacc393d103063234e8bc28"
+  },
+  "schema_version": 1,
+  "shared_query_ids_sha256": "0f1a717432f0d6fe9c77931d4e4ca312116d7a6641e4506afd52f73c3abf56a4",
+  "smoke": false,
+  "training": {
+    "effective_batch_size": 16,
+    "fresh_rollout_exposures": 13776,
+    "gradient_checkpointing": true,
+    "gradient_checkpointing_use_reentrant": false,
+    "gradient_clip": 1.0,
+    "learning_rate": 1e-05,
+    "lora_alpha": 32,
+    "lora_dropout": 0.0,
+    "lora_rank": 16,
+    "microbatch_size": 1,
+    "objective": "full_vocabulary_forward_kl_teacher_to_student",
+    "optimizer": "AdamW",
+    "optimizer_steps": 861,
+    "precision": "bfloat16",
+    "scheduler": "constant",
+    "target_modules": [
+      "q_proj",
+      "k_proj",
+      "v_proj",
+      "o_proj",
+      "gate_proj",
+      "up_proj",
+      "down_proj"
+    ],
+    "weight_decay": 0.01
+  }
+}
diff --git a/adapters/big_math-demonstration-external/tokenizer.json b/adapters/big_math-demonstration-external/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-demonstration-external/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-demonstration-external/tokenizer_config.json b/adapters/big_math-demonstration-external/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..33134953c69ef0d8f6d87a813851b965fdbe84bf
--- /dev/null
+++ b/adapters/big_math-demonstration-external/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/adapters/big_math-demonstration-external/training_complete.json b/adapters/big_math-demonstration-external/training_complete.json
new file mode 100644
index 0000000000000000000000000000000000000000..c99840326bb72d037d14583fe22f16fd01c4b11a
--- /dev/null
+++ b/adapters/big_math-demonstration-external/training_complete.json
@@ -0,0 +1,14 @@
+{
+  "capped_rollouts": 5,
+  "condition": "big_math-demonstration-external",
+  "final_adapter": "artifacts/opd/training/big_math-demonstration-external/final",
+  "finite_loss_verified": true,
+  "fresh_rollout_exposures": 13776,
+  "last_loss": 0.12790197663349787,
+  "manifest_sha256": "747180dc8747c47ada51ac721676b06794012bcb2588a0136baf9a9ff380b8d9",
+  "resolved_config_sha256": "bc3cad94dbd82970520e3c1c2964799c91b7550a7457f076b46f0aae2ce0ed6a",
+  "schema_version": 1,
+  "smoke": false,
+  "steps": 861,
+  "teacher_frozen_verified": true
+}
diff --git a/adapters/big_math-demonstration-self_supervised/adapter_model.safetensors b/adapters/big_math-demonstration-self_supervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..b7179203400314a6657aa11131fe72db41d9febd
--- /dev/null
+++ b/adapters/big_math-demonstration-self_supervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4f84a0bd8474d7025ef04f43f7d50c3a16c84dd4beb57f0c1b0a9cec4c118ce
+size 161533192
diff --git a/adapters/big_math-demonstration-self_supervised/tokenizer.json b/adapters/big_math-demonstration-self_supervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-demonstration-self_supervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-demonstration-self_unsupervised/adapter_model.safetensors b/adapters/big_math-demonstration-self_unsupervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..1936a0a681babc4f653001643b24930ca4433816
--- /dev/null
+++ b/adapters/big_math-demonstration-self_unsupervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15fe1edd2f7a2ed108ad17244643b888786aa38be637b4d538612e6c859a07d8
+size 161533192
diff --git a/adapters/big_math-demonstration-self_unsupervised/tokenizer.json b/adapters/big_math-demonstration-self_unsupervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-demonstration-self_unsupervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-feedback-external/README.md b/adapters/big_math-feedback-external/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..eb09148de648cee968a2971cee1f2dc513f83603
--- /dev/null
+++ b/adapters/big_math-feedback-external/README.md
@@ -0,0 +1,207 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **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]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### 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
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/adapters/big_math-feedback-external/adapter_config.json b/adapters/big_math-feedback-external/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..447fc94bb2d1fbac389cc7c09b2b0b70db90657c
--- /dev/null
+++ b/adapters/big_math-feedback-external/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
+  "target_modules": [
+    "gate_proj",
+    "k_proj",
+    "down_proj",
+    "o_proj",
+    "up_proj",
+    "q_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
diff --git a/adapters/big_math-feedback-external/adapter_model.safetensors b/adapters/big_math-feedback-external/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..e9996dd9c049270ff2a8bcac0896a80b6021da65
--- /dev/null
+++ b/adapters/big_math-feedback-external/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b5f4cfe68c17fbb778c88b20ce4fec1596f3abd6eace7c320f445edf99c1576
+size 161533192
diff --git a/adapters/big_math-feedback-external/chat_template.jinja b/adapters/big_math-feedback-external/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e
--- /dev/null
+++ b/adapters/big_math-feedback-external/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/adapters/big_math-feedback-external/resolved_training_config.json b/adapters/big_math-feedback-external/resolved_training_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..80eb27daa9ce5f3d95eda8f5386b081e8beeb22c
--- /dev/null
+++ b/adapters/big_math-feedback-external/resolved_training_config.json
@@ -0,0 +1,39 @@
+{
+  "condition": "big_math-feedback-external",
+  "context_manifest_sha256": "e1e4ae68ffbfcfa4b0e58f889524fdffd44fc047cf7cd7b5ea51eb0443fbfffa",
+  "manifest_sha256": "8bbf8156f88d49ec7ef8cb206f214a2cd7c08146a8bb9c03d581231f4c4bed13",
+  "model": {
+    "repo": "Qwen/Qwen2.5-7B-Instruct",
+    "revision": "a09a35458c702b33eeacc393d103063234e8bc28"
+  },
+  "schema_version": 1,
+  "shared_query_ids_sha256": "0f1a717432f0d6fe9c77931d4e4ca312116d7a6641e4506afd52f73c3abf56a4",
+  "smoke": false,
+  "training": {
+    "effective_batch_size": 16,
+    "fresh_rollout_exposures": 13776,
+    "gradient_checkpointing": true,
+    "gradient_checkpointing_use_reentrant": false,
+    "gradient_clip": 1.0,
+    "learning_rate": 1e-05,
+    "lora_alpha": 32,
+    "lora_dropout": 0.0,
+    "lora_rank": 16,
+    "microbatch_size": 1,
+    "objective": "full_vocabulary_forward_kl_teacher_to_student",
+    "optimizer": "AdamW",
+    "optimizer_steps": 861,
+    "precision": "bfloat16",
+    "scheduler": "constant",
+    "target_modules": [
+      "q_proj",
+      "k_proj",
+      "v_proj",
+      "o_proj",
+      "gate_proj",
+      "up_proj",
+      "down_proj"
+    ],
+    "weight_decay": 0.01
+  }
+}
diff --git a/adapters/big_math-feedback-external/tokenizer.json b/adapters/big_math-feedback-external/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-feedback-external/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-feedback-external/tokenizer_config.json b/adapters/big_math-feedback-external/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..33134953c69ef0d8f6d87a813851b965fdbe84bf
--- /dev/null
+++ b/adapters/big_math-feedback-external/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/adapters/big_math-feedback-external/training_complete.json b/adapters/big_math-feedback-external/training_complete.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d7ce1343594880050cd5a9344a49e3f7c2eb0c1
--- /dev/null
+++ b/adapters/big_math-feedback-external/training_complete.json
@@ -0,0 +1,14 @@
+{
+  "capped_rollouts": 6,
+  "condition": "big_math-feedback-external",
+  "final_adapter": "artifacts/opd/training/big_math-feedback-external/final",
+  "finite_loss_verified": true,
+  "fresh_rollout_exposures": 13776,
+  "last_loss": 0.073949893768109,
+  "manifest_sha256": "3d63c3fef07302068925ab5c3079552c1df45bac1428cf1ccfd33b2c9e6500ea",
+  "resolved_config_sha256": "8bbf8156f88d49ec7ef8cb206f214a2cd7c08146a8bb9c03d581231f4c4bed13",
+  "schema_version": 1,
+  "smoke": false,
+  "steps": 861,
+  "teacher_frozen_verified": true
+}
diff --git a/adapters/big_math-feedback-self_supervised/README.md b/adapters/big_math-feedback-self_supervised/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..eb09148de648cee968a2971cee1f2dc513f83603
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/README.md
@@ -0,0 +1,207 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **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]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### 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
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/adapters/big_math-feedback-self_supervised/adapter_config.json b/adapters/big_math-feedback-self_supervised/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..47123ef710bfcc3960d97daae495c7ce2375e5d2
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
+  "target_modules": [
+    "q_proj",
+    "v_proj",
+    "down_proj",
+    "gate_proj",
+    "k_proj",
+    "o_proj",
+    "up_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
diff --git a/adapters/big_math-feedback-self_supervised/adapter_model.safetensors b/adapters/big_math-feedback-self_supervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d678568ef8d8857cc6e004352f327e66356992fe
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b75b1f174f99fdaaa1bc7cc0b6542918b13e1dfeb7e3a3e9d4792e79ba1708f
+size 161533192
diff --git a/adapters/big_math-feedback-self_supervised/chat_template.jinja b/adapters/big_math-feedback-self_supervised/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/adapters/big_math-feedback-self_supervised/resolved_training_config.json b/adapters/big_math-feedback-self_supervised/resolved_training_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..65b9c8ecad96994d4f250fbbeefea78a8669e5a2
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/resolved_training_config.json
@@ -0,0 +1,39 @@
+{
+  "condition": "big_math-feedback-self_supervised",
+  "context_manifest_sha256": "1ddfabc585c01cb7db748cacb93240142ca53acfe996db15251bc2c64bb208f3",
+  "manifest_sha256": "df9804754787fe0b85eac459378af44180badf64bdbf9b256ca2adbfac634d9c",
+  "model": {
+    "repo": "Qwen/Qwen2.5-7B-Instruct",
+    "revision": "a09a35458c702b33eeacc393d103063234e8bc28"
+  },
+  "schema_version": 1,
+  "shared_query_ids_sha256": "0f1a717432f0d6fe9c77931d4e4ca312116d7a6641e4506afd52f73c3abf56a4",
+  "smoke": false,
+  "training": {
+    "effective_batch_size": 16,
+    "fresh_rollout_exposures": 13776,
+    "gradient_checkpointing": true,
+    "gradient_checkpointing_use_reentrant": false,
+    "gradient_clip": 1.0,
+    "learning_rate": 1e-05,
+    "lora_alpha": 32,
+    "lora_dropout": 0.0,
+    "lora_rank": 16,
+    "microbatch_size": 1,
+    "objective": "full_vocabulary_forward_kl_teacher_to_student",
+    "optimizer": "AdamW",
+    "optimizer_steps": 861,
+    "precision": "bfloat16",
+    "scheduler": "constant",
+    "target_modules": [
+      "q_proj",
+      "k_proj",
+      "v_proj",
+      "o_proj",
+      "gate_proj",
+      "up_proj",
+      "down_proj"
+    ],
+    "weight_decay": 0.01
+  }
+}
diff --git a/adapters/big_math-feedback-self_supervised/tokenizer.json b/adapters/big_math-feedback-self_supervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-feedback-self_supervised/tokenizer_config.json b/adapters/big_math-feedback-self_supervised/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..33134953c69ef0d8f6d87a813851b965fdbe84bf
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/adapters/big_math-feedback-self_supervised/training_complete.json b/adapters/big_math-feedback-self_supervised/training_complete.json
new file mode 100644
index 0000000000000000000000000000000000000000..79fc66dfe1cff3ad2947b4122e7b06baa83caf1d
--- /dev/null
+++ b/adapters/big_math-feedback-self_supervised/training_complete.json
@@ -0,0 +1,14 @@
+{
+  "capped_rollouts": 2,
+  "condition": "big_math-feedback-self_supervised",
+  "final_adapter": "artifacts/opd/training/big_math-feedback-self_supervised/final",
+  "finite_loss_verified": true,
+  "fresh_rollout_exposures": 13776,
+  "last_loss": 0.07367482130622557,
+  "manifest_sha256": "f9f618b1276cd5b57084470ddb66d0350b237b0fe9fc00d6cbb8d18e29a556bd",
+  "resolved_config_sha256": "df9804754787fe0b85eac459378af44180badf64bdbf9b256ca2adbfac634d9c",
+  "schema_version": 1,
+  "smoke": false,
+  "steps": 861,
+  "teacher_frozen_verified": true
+}
diff --git a/adapters/big_math-feedback-self_unsupervised/README.md b/adapters/big_math-feedback-self_unsupervised/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..eb09148de648cee968a2971cee1f2dc513f83603
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/README.md
@@ -0,0 +1,207 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **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]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### 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
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/adapters/big_math-feedback-self_unsupervised/adapter_config.json b/adapters/big_math-feedback-self_unsupervised/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c49729270e742aa5883e58b2df0c2f5a55765547
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
+  "target_modules": [
+    "gate_proj",
+    "v_proj",
+    "down_proj",
+    "q_proj",
+    "o_proj",
+    "up_proj",
+    "k_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
diff --git a/adapters/big_math-feedback-self_unsupervised/adapter_model.safetensors b/adapters/big_math-feedback-self_unsupervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..55f4e300d787baa170aac758d7e63885058d2f8a
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f66071b4c3cf5da15b1a742f1428de7e3664223d6538b7750cb0ade2cfecad17
+size 161533192
diff --git a/adapters/big_math-feedback-self_unsupervised/chat_template.jinja b/adapters/big_math-feedback-self_unsupervised/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/adapters/big_math-feedback-self_unsupervised/resolved_training_config.json b/adapters/big_math-feedback-self_unsupervised/resolved_training_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac2775120c32c35051651ebdb6f6896a40b872e4
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/resolved_training_config.json
@@ -0,0 +1,39 @@
+{
+  "condition": "big_math-feedback-self_unsupervised",
+  "context_manifest_sha256": "31718cd7880de4fbdc51d554c48ae440315444e7451ccbde54c029e410e83dad",
+  "manifest_sha256": "2dd25c1f7eabe96510b9ab0417d4e18883052f082fabb0b1e07f589a82be9563",
+  "model": {
+    "repo": "Qwen/Qwen2.5-7B-Instruct",
+    "revision": "a09a35458c702b33eeacc393d103063234e8bc28"
+  },
+  "schema_version": 1,
+  "shared_query_ids_sha256": "0f1a717432f0d6fe9c77931d4e4ca312116d7a6641e4506afd52f73c3abf56a4",
+  "smoke": false,
+  "training": {
+    "effective_batch_size": 16,
+    "fresh_rollout_exposures": 13776,
+    "gradient_checkpointing": true,
+    "gradient_checkpointing_use_reentrant": false,
+    "gradient_clip": 1.0,
+    "learning_rate": 1e-05,
+    "lora_alpha": 32,
+    "lora_dropout": 0.0,
+    "lora_rank": 16,
+    "microbatch_size": 1,
+    "objective": "full_vocabulary_forward_kl_teacher_to_student",
+    "optimizer": "AdamW",
+    "optimizer_steps": 861,
+    "precision": "bfloat16",
+    "scheduler": "constant",
+    "target_modules": [
+      "q_proj",
+      "k_proj",
+      "v_proj",
+      "o_proj",
+      "gate_proj",
+      "up_proj",
+      "down_proj"
+    ],
+    "weight_decay": 0.01
+  }
+}
diff --git a/adapters/big_math-feedback-self_unsupervised/tokenizer.json b/adapters/big_math-feedback-self_unsupervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-feedback-self_unsupervised/tokenizer_config.json b/adapters/big_math-feedback-self_unsupervised/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..33134953c69ef0d8f6d87a813851b965fdbe84bf
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/adapters/big_math-feedback-self_unsupervised/training_complete.json b/adapters/big_math-feedback-self_unsupervised/training_complete.json
new file mode 100644
index 0000000000000000000000000000000000000000..ffe4fadc120d53910fe382211b0ee88c3eb8f8a2
--- /dev/null
+++ b/adapters/big_math-feedback-self_unsupervised/training_complete.json
@@ -0,0 +1,14 @@
+{
+  "capped_rollouts": 5,
+  "condition": "big_math-feedback-self_unsupervised",
+  "final_adapter": "artifacts/opd/training/big_math-feedback-self_unsupervised/final",
+  "finite_loss_verified": true,
+  "fresh_rollout_exposures": 13776,
+  "last_loss": 0.07397536791435096,
+  "manifest_sha256": "1e493901ba6b71be2792244a94674f2f264e74339d2f7fa41d2600121bc7875e",
+  "resolved_config_sha256": "2dd25c1f7eabe96510b9ab0417d4e18883052f082fabb0b1e07f589a82be9563",
+  "schema_version": 1,
+  "smoke": false,
+  "steps": 861,
+  "teacher_frozen_verified": true
+}
diff --git a/adapters/big_math-rephrase-external/adapter_model.safetensors b/adapters/big_math-rephrase-external/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..eeb70c9be1f7b2003982ba3f609aa9387c44706a
--- /dev/null
+++ b/adapters/big_math-rephrase-external/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7879963139a1dc0d33e8bbd016a5231dbf50a7ed95fc307b646ff7398fb8bbec
+size 161533192
diff --git a/adapters/big_math-rephrase-external/tokenizer.json b/adapters/big_math-rephrase-external/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-rephrase-external/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-rephrase-self_supervised/README.md b/adapters/big_math-rephrase-self_supervised/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..eb09148de648cee968a2971cee1f2dc513f83603
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/README.md
@@ -0,0 +1,207 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **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]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### 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
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/adapters/big_math-rephrase-self_supervised/adapter_config.json b/adapters/big_math-rephrase-self_supervised/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..140a7d82853a4dbb1314573f3191e533a1e3c4ba
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
+  "target_modules": [
+    "o_proj",
+    "k_proj",
+    "down_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
diff --git a/adapters/big_math-rephrase-self_supervised/adapter_model.safetensors b/adapters/big_math-rephrase-self_supervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..1e36a625f38c661d8dd46251745e82f5a841bf71
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15c10060693bc7d0f07d7f54b94122a9076c7da1776818a4319d0cb2ae66a81f
+size 161533192
diff --git a/adapters/big_math-rephrase-self_supervised/chat_template.jinja b/adapters/big_math-rephrase-self_supervised/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/adapters/big_math-rephrase-self_supervised/resolved_training_config.json b/adapters/big_math-rephrase-self_supervised/resolved_training_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..36061f5ff0dde07e612fd97ae475bb7c73c3d609
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/resolved_training_config.json
@@ -0,0 +1,39 @@
+{
+  "condition": "big_math-rephrase-self_supervised",
+  "context_manifest_sha256": "476a554bdbec03fa2b1f47529a97a1d6ce7dbe52ea716d4e35f4490be6e0b29d",
+  "manifest_sha256": "685c13a4c7627ca8ba5519c3bef3ccbb8335643764f4a04f68311278f7d64d59",
+  "model": {
+    "repo": "Qwen/Qwen2.5-7B-Instruct",
+    "revision": "a09a35458c702b33eeacc393d103063234e8bc28"
+  },
+  "schema_version": 1,
+  "shared_query_ids_sha256": "0f1a717432f0d6fe9c77931d4e4ca312116d7a6641e4506afd52f73c3abf56a4",
+  "smoke": false,
+  "training": {
+    "effective_batch_size": 16,
+    "fresh_rollout_exposures": 13776,
+    "gradient_checkpointing": true,
+    "gradient_checkpointing_use_reentrant": false,
+    "gradient_clip": 1.0,
+    "learning_rate": 1e-05,
+    "lora_alpha": 32,
+    "lora_dropout": 0.0,
+    "lora_rank": 16,
+    "microbatch_size": 1,
+    "objective": "full_vocabulary_forward_kl_teacher_to_student",
+    "optimizer": "AdamW",
+    "optimizer_steps": 861,
+    "precision": "bfloat16",
+    "scheduler": "constant",
+    "target_modules": [
+      "q_proj",
+      "k_proj",
+      "v_proj",
+      "o_proj",
+      "gate_proj",
+      "up_proj",
+      "down_proj"
+    ],
+    "weight_decay": 0.01
+  }
+}
diff --git a/adapters/big_math-rephrase-self_supervised/tokenizer.json b/adapters/big_math-rephrase-self_supervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/big_math-rephrase-self_supervised/tokenizer_config.json b/adapters/big_math-rephrase-self_supervised/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..33134953c69ef0d8f6d87a813851b965fdbe84bf
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/adapters/big_math-rephrase-self_supervised/training_complete.json b/adapters/big_math-rephrase-self_supervised/training_complete.json
new file mode 100644
index 0000000000000000000000000000000000000000..4dfd285e4e4f346afb92742f50e6559fad505588
--- /dev/null
+++ b/adapters/big_math-rephrase-self_supervised/training_complete.json
@@ -0,0 +1,14 @@
+{
+  "capped_rollouts": 12,
+  "condition": "big_math-rephrase-self_supervised",
+  "final_adapter": "artifacts/opd/training/big_math-rephrase-self_supervised/final",
+  "finite_loss_verified": true,
+  "fresh_rollout_exposures": 13776,
+  "last_loss": 0.005049956341584523,
+  "manifest_sha256": "0fe96d488f86432fff3dda8a1b60d4c195ad6e1d8ec7d0e95ec288ccb146c0d4",
+  "resolved_config_sha256": "685c13a4c7627ca8ba5519c3bef3ccbb8335643764f4a04f68311278f7d64d59",
+  "schema_version": 1,
+  "smoke": false,
+  "steps": 861,
+  "teacher_frozen_verified": true
+}
diff --git a/adapters/big_math-rephrase-self_unsupervised/adapter_model.safetensors b/adapters/big_math-rephrase-self_unsupervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c6f0fc5d3fee411696ddfd9659e9fd1a03d573c8
--- /dev/null
+++ b/adapters/big_math-rephrase-self_unsupervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d3bee5e41be07bdd0a060796eaeda241ae3a094e3871739d0ab9257b9b62632
+size 161533192
diff --git a/adapters/big_math-rephrase-self_unsupervised/tokenizer.json b/adapters/big_math-rephrase-self_unsupervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/big_math-rephrase-self_unsupervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/proofwriter-feedback-external/adapter_model.safetensors b/adapters/proofwriter-feedback-external/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..7e6afbd842c4d781d89ba3b5a0c89e171c026d33
--- /dev/null
+++ b/adapters/proofwriter-feedback-external/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e091c058784b0c0e046d29ae93024e49d5c45d4c64626806a949dc12e69ba131
+size 161533192
diff --git a/adapters/proofwriter-feedback-external/tokenizer.json b/adapters/proofwriter-feedback-external/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/proofwriter-feedback-external/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/proofwriter-feedback-self_supervised/adapter_model.safetensors b/adapters/proofwriter-feedback-self_supervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..bfcacf2020789e679c64697264b6a0de709ee9d1
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_supervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10ea6ae38582bee68e3b64d14af43e5eadad095f558b7cd9f5db2765fda07732
+size 161533192
diff --git a/adapters/proofwriter-feedback-self_supervised/tokenizer.json b/adapters/proofwriter-feedback-self_supervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_supervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/proofwriter-feedback-self_unsupervised/README.md b/adapters/proofwriter-feedback-self_unsupervised/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..eb09148de648cee968a2971cee1f2dc513f83603
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/README.md
@@ -0,0 +1,207 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **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]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### 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
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/adapters/proofwriter-feedback-self_unsupervised/adapter_config.json b/adapters/proofwriter-feedback-self_unsupervised/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..849ec661b82acbe8ed61d7340828f49a1181b849
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
+  "target_modules": [
+    "k_proj",
+    "v_proj",
+    "up_proj",
+    "down_proj",
+    "o_proj",
+    "gate_proj",
+    "q_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
diff --git a/adapters/proofwriter-feedback-self_unsupervised/adapter_model.safetensors b/adapters/proofwriter-feedback-self_unsupervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..4ad4fadec78c8d842a1eb2173347a46065945792
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:358906f9ef26c53cd3995dfdd42fdfd428149e9a13642b7398dc46313ca7c9db
+size 161533192
diff --git a/adapters/proofwriter-feedback-self_unsupervised/chat_template.jinja b/adapters/proofwriter-feedback-self_unsupervised/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/adapters/proofwriter-feedback-self_unsupervised/resolved_training_config.json b/adapters/proofwriter-feedback-self_unsupervised/resolved_training_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..6136f164282e780ff75c123e52a920f82e6bcc13
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/resolved_training_config.json
@@ -0,0 +1,39 @@
+{
+  "condition": "proofwriter-feedback-self_unsupervised",
+  "context_manifest_sha256": "477a7dfbff954619f1a9591ea7c432abb9acf8eab6147107165eb0b4fbd33c12",
+  "manifest_sha256": "dc36923e42a3f77f534509530bdb6f99891900575ffb36d8794b4efc1e54a97b",
+  "model": {
+    "repo": "Qwen/Qwen2.5-7B-Instruct",
+    "revision": "a09a35458c702b33eeacc393d103063234e8bc28"
+  },
+  "schema_version": 1,
+  "shared_query_ids_sha256": "936371eac5bc59d4bd20130ce46419c78a18f1a93a27cb1bd3b3dd5f386a1ad9",
+  "smoke": false,
+  "training": {
+    "effective_batch_size": 16,
+    "fresh_rollout_exposures": 13776,
+    "gradient_checkpointing": true,
+    "gradient_checkpointing_use_reentrant": false,
+    "gradient_clip": 1.0,
+    "learning_rate": 1e-05,
+    "lora_alpha": 32,
+    "lora_dropout": 0.0,
+    "lora_rank": 16,
+    "microbatch_size": 1,
+    "objective": "full_vocabulary_forward_kl_teacher_to_student",
+    "optimizer": "AdamW",
+    "optimizer_steps": 861,
+    "precision": "bfloat16",
+    "scheduler": "constant",
+    "target_modules": [
+      "q_proj",
+      "k_proj",
+      "v_proj",
+      "o_proj",
+      "gate_proj",
+      "up_proj",
+      "down_proj"
+    ],
+    "weight_decay": 0.01
+  }
+}
diff --git a/adapters/proofwriter-feedback-self_unsupervised/tokenizer.json b/adapters/proofwriter-feedback-self_unsupervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/proofwriter-feedback-self_unsupervised/tokenizer_config.json b/adapters/proofwriter-feedback-self_unsupervised/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..33134953c69ef0d8f6d87a813851b965fdbe84bf
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/adapters/proofwriter-feedback-self_unsupervised/training_complete.json b/adapters/proofwriter-feedback-self_unsupervised/training_complete.json
new file mode 100644
index 0000000000000000000000000000000000000000..26761a425d2bfef3d787386fea15bc4220550d58
--- /dev/null
+++ b/adapters/proofwriter-feedback-self_unsupervised/training_complete.json
@@ -0,0 +1,14 @@
+{
+  "capped_rollouts": 1,
+  "condition": "proofwriter-feedback-self_unsupervised",
+  "final_adapter": "artifacts/opd/training/proofwriter-feedback-self_unsupervised/final",
+  "finite_loss_verified": true,
+  "fresh_rollout_exposures": 13776,
+  "last_loss": 0.1537594004706058,
+  "manifest_sha256": "ed2d14b91233b61080276301d803e53ba7fd1c5d3f91de5f4751a0a713f5ffc5",
+  "resolved_config_sha256": "dc36923e42a3f77f534509530bdb6f99891900575ffb36d8794b4efc1e54a97b",
+  "schema_version": 1,
+  "smoke": false,
+  "steps": 861,
+  "teacher_frozen_verified": true
+}
diff --git a/adapters/proofwriter-rephrase-external/adapter_model.safetensors b/adapters/proofwriter-rephrase-external/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..06e282540d0b7a58e09580ec629e2405acbba323
--- /dev/null
+++ b/adapters/proofwriter-rephrase-external/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df92961d1b9de0ef45fb49dad03679c5f1bea2e6e1dd95e97d5258664ba07905
+size 161533192
diff --git a/adapters/proofwriter-rephrase-external/tokenizer.json b/adapters/proofwriter-rephrase-external/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/proofwriter-rephrase-external/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/proofwriter-rephrase-self_supervised/adapter_model.safetensors b/adapters/proofwriter-rephrase-self_supervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d132b8c8c02c94608b5d0172b266557d9485815c
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_supervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a5299a718c102078c11a233a014bab164976a77b2b1ff4706a592b0a577d5c0
+size 161533192
diff --git a/adapters/proofwriter-rephrase-self_supervised/tokenizer.json b/adapters/proofwriter-rephrase-self_supervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_supervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/README.md b/adapters/proofwriter-rephrase-self_unsupervised/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..eb09148de648cee968a2971cee1f2dc513f83603
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/README.md
@@ -0,0 +1,207 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **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]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### 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
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/adapter_config.json b/adapters/proofwriter-rephrase-self_unsupervised/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..933234fb0181330a116a6ea29559babdb68384a0
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
+  "target_modules": [
+    "o_proj",
+    "q_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj",
+    "v_proj",
+    "k_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/adapter_model.safetensors b/adapters/proofwriter-rephrase-self_unsupervised/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..a8c4109f472135754184ff4a21975ebba48d3071
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55222c3d5dcfbca8aea9ddf4dafcbeff8ae80717ea8e866a808d40a4c3f90697
+size 161533192
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/chat_template.jinja b/adapters/proofwriter-rephrase-self_unsupervised/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/resolved_training_config.json b/adapters/proofwriter-rephrase-self_unsupervised/resolved_training_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..32511a00af480c20ba4b1daf9f62f9bc74ecacd6
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/resolved_training_config.json
@@ -0,0 +1,39 @@
+{
+  "condition": "proofwriter-rephrase-self_unsupervised",
+  "context_manifest_sha256": "61c1401e9e0326b6e30412d594015cdb1a6962e826d943435613a685600d1595",
+  "manifest_sha256": "5523743a71aa023db88906801624159a6c4e002c44a0b5043d89b02fc83850c9",
+  "model": {
+    "repo": "Qwen/Qwen2.5-7B-Instruct",
+    "revision": "a09a35458c702b33eeacc393d103063234e8bc28"
+  },
+  "schema_version": 1,
+  "shared_query_ids_sha256": "936371eac5bc59d4bd20130ce46419c78a18f1a93a27cb1bd3b3dd5f386a1ad9",
+  "smoke": false,
+  "training": {
+    "effective_batch_size": 16,
+    "fresh_rollout_exposures": 13776,
+    "gradient_checkpointing": true,
+    "gradient_checkpointing_use_reentrant": false,
+    "gradient_clip": 1.0,
+    "learning_rate": 1e-05,
+    "lora_alpha": 32,
+    "lora_dropout": 0.0,
+    "lora_rank": 16,
+    "microbatch_size": 1,
+    "objective": "full_vocabulary_forward_kl_teacher_to_student",
+    "optimizer": "AdamW",
+    "optimizer_steps": 861,
+    "precision": "bfloat16",
+    "scheduler": "constant",
+    "target_modules": [
+      "q_proj",
+      "k_proj",
+      "v_proj",
+      "o_proj",
+      "gate_proj",
+      "up_proj",
+      "down_proj"
+    ],
+    "weight_decay": 0.01
+  }
+}
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/tokenizer.json b/adapters/proofwriter-rephrase-self_unsupervised/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/tokenizer_config.json b/adapters/proofwriter-rephrase-self_unsupervised/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..33134953c69ef0d8f6d87a813851b965fdbe84bf
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
diff --git a/adapters/proofwriter-rephrase-self_unsupervised/training_complete.json b/adapters/proofwriter-rephrase-self_unsupervised/training_complete.json
new file mode 100644
index 0000000000000000000000000000000000000000..3191a96eba8ef8b5247b1b0f647ea8580c0b70c5
--- /dev/null
+++ b/adapters/proofwriter-rephrase-self_unsupervised/training_complete.json
@@ -0,0 +1,14 @@
+{
+  "capped_rollouts": 5,
+  "condition": "proofwriter-rephrase-self_unsupervised",
+  "final_adapter": "artifacts/opd/training/proofwriter-rephrase-self_unsupervised/final",
+  "finite_loss_verified": true,
+  "fresh_rollout_exposures": 13776,
+  "last_loss": 0.017555344909910234,
+  "manifest_sha256": "aa5e08b3aecc38b0f80816075fbe2767ed6242ce0feeaac47d665a61e370b475",
+  "resolved_config_sha256": "5523743a71aa023db88906801624159a6c4e002c44a0b5043d89b02fc83850c9",
+  "schema_version": 1,
+  "smoke": false,
+  "steps": 861,
+  "teacher_frozen_verified": true
+}