Add files using upload-large-folder tool

cls_gemma_baseline/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_baseline/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_baseline/checkpoint-200/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_baseline/checkpoint-200/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_baseline/checkpoint-200/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_baseline/checkpoint-400/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_baseline/checkpoint-400/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_baseline/checkpoint-400/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_baseline/train_metrics.json

@@ -0,0 +1,1874 @@
+{
+  "loss_log": [
+    {
+      "global_step": 20,
+      "optimizer_step": 2,
+      "lr": 4.347826086956522e-06,
+      "lm_loss": 3.1184232234954834,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 3.1184232234954834
+    },
+    {
+      "global_step": 40,
+      "optimizer_step": 5,
+      "lr": 1.0869565217391305e-05,
+      "lm_loss": 1.0916705131530762,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 1.0916705131530762
+    },
+    {
+      "global_step": 60,
+      "optimizer_step": 7,
+      "lr": 1.5217391304347828e-05,
+      "lm_loss": 0.6041339039802551,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.6041339039802551
+    },
+    {
+      "global_step": 80,
+      "optimizer_step": 10,
+      "lr": 2.173913043478261e-05,
+      "lm_loss": 1.8147910833358765,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 1.8147910833358765
+    },
+    {
+      "global_step": 100,
+      "optimizer_step": 12,
+      "lr": 2.608695652173913e-05,
+      "lm_loss": 0.5634512305259705,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.5634512305259705
+    },
+    {
+      "global_step": 120,
+      "optimizer_step": 15,
+      "lr": 3.260869565217392e-05,
+      "lm_loss": 2.721200942993164,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 2.721200942993164
+    },
+    {
+      "global_step": 140,
+      "optimizer_step": 17,
+      "lr": 3.695652173913043e-05,
+      "lm_loss": 0.6781831979751587,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.6781831979751587
+    },
+    {
+      "global_step": 160,
+      "optimizer_step": 20,
+      "lr": 4.347826086956522e-05,
+      "lm_loss": 0.41744333505630493,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.41744333505630493
+    },
+    {
+      "global_step": 180,
+      "optimizer_step": 22,
+      "lr": 4.782608695652174e-05,
+      "lm_loss": 0.06321373581886292,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.06321373581886292
+    },
+    {
+      "global_step": 200,
+      "optimizer_step": 25,
+      "lr": 4.9997541293989384e-05,
+      "lm_loss": 0.02435319498181343,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.02435319498181343
+    },
+    {
+      "global_step": 220,
+      "optimizer_step": 27,
+      "lr": 4.999016565957633e-05,
+      "lm_loss": 0.019743939861655235,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.019743939861655235
+    },
+    {
+      "global_step": 240,
+      "optimizer_step": 30,
+      "lr": 4.996988640512931e-05,
+      "lm_loss": 0.3459993302822113,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.3459993302822113
+    },
+    {
+      "global_step": 260,
+      "optimizer_step": 32,
+      "lr": 4.995022691166621e-05,
+      "lm_loss": 0.0298320259898901,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0298320259898901
+    },
+    {
+      "global_step": 280,
+      "optimizer_step": 35,
+      "lr": 4.991153735294049e-05,
+      "lm_loss": 0.46130573749542236,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.46130573749542236
+    },
+    {
+      "global_step": 300,
+      "optimizer_step": 37,
+      "lr": 4.987961816680492e-05,
+      "lm_loss": 0.1003972664475441,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1003972664475441
+    },
+    {
+      "global_step": 320,
+      "optimizer_step": 40,
+      "lr": 4.9822565862832806e-05,
+      "lm_loss": 0.13187472522258759,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.13187472522258759
+    },
+    {
+      "global_step": 340,
+      "optimizer_step": 42,
+      "lr": 4.9778426220595456e-05,
+      "lm_loss": 0.32335221767425537,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.32335221767425537
+    },
+    {
+      "global_step": 360,
+      "optimizer_step": 45,
+      "lr": 4.970308130276272e-05,
+      "lm_loss": 0.7211238741874695,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.7211238741874695
+    },
+    {
+      "global_step": 380,
+      "optimizer_step": 47,
+      "lr": 4.96467754629559e-05,
+      "lm_loss": 0.03598100692033768,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.03598100692033768
+    },
+    {
+      "global_step": 400,
+      "optimizer_step": 50,
+      "lr": 4.955323054879298e-05,
+      "lm_loss": 0.10630018264055252,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.10630018264055252
+    },
+    {
+      "global_step": 420,
+      "optimizer_step": 52,
+      "lr": 4.948482772521345e-05,
+      "lm_loss": 0.11145680397748947,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.11145680397748947
+    },
+    {
+      "global_step": 440,
+      "optimizer_step": 55,
+      "lr": 4.937319780454559e-05,
+      "lm_loss": 0.18087011575698853,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.18087011575698853
+    },
+    {
+      "global_step": 460,
+      "optimizer_step": 57,
+      "lr": 4.929278208117378e-05,
+      "lm_loss": 0.6563955545425415,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.6563955545425415
+    },
+    {
+      "global_step": 480,
+      "optimizer_step": 60,
+      "lr": 4.916320437477008e-05,
+      "lm_loss": 0.32431942224502563,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.32431942224502563
+    },
+    {
+      "global_step": 500,
+      "optimizer_step": 62,
+      "lr": 4.907087460240999e-05,
+      "lm_loss": 0.14991332590579987,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.14991332590579987
+    },
+    {
+      "global_step": 520,
+      "optimizer_step": 65,
+      "lr": 4.892350839330522e-05,
+      "lm_loss": 0.1010974645614624,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1010974645614624
+    },
+    {
+      "global_step": 540,
+      "optimizer_step": 67,
+      "lr": 4.881937806807241e-05,
+      "lm_loss": 0.15496166050434113,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.15496166050434113
+    },
+    {
+      "global_step": 560,
+      "optimizer_step": 70,
+      "lr": 4.865440450576874e-05,
+      "lm_loss": 0.026232725009322166,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.026232725009322166
+    },
+    {
+      "global_step": 580,
+      "optimizer_step": 72,
+      "lr": 4.853860162957552e-05,
+      "lm_loss": 0.09959332644939423,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.09959332644939423
+    },
+    {
+      "global_step": 600,
+      "optimizer_step": 75,
+      "lr": 4.8356223507364996e-05,
+      "lm_loss": 0.01794126257300377,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.01794126257300377
+    },
+    {
+      "global_step": 620,
+      "optimizer_step": 77,
+      "lr": 4.822889043057446e-05,
+      "lm_loss": 0.26736268401145935,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.26736268401145935
+    },
+    {
+      "global_step": 640,
+      "optimizer_step": 80,
+      "lr": 4.802933193625593e-05,
+      "lm_loss": 0.007746524177491665,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.007746524177491665
+    },
+    {
+      "global_step": 660,
+      "optimizer_step": 82,
+      "lr": 4.789062518269812e-05,
+      "lm_loss": 0.03736429661512375,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.03736429661512375
+    },
+    {
+      "global_step": 680,
+      "optimizer_step": 85,
+      "lr": 4.767413162299501e-05,
+      "lm_loss": 0.06276468187570572,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.06276468187570572
+    },
+    {
+      "global_step": 700,
+      "optimizer_step": 87,
+      "lr": 4.752422169756048e-05,
+      "lm_loss": 0.22046461701393127,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.22046461701393127
+    },
+    {
+      "global_step": 720,
+      "optimizer_step": 90,
+      "lr": 4.729105919657818e-05,
+      "lm_loss": 0.4013015925884247,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.4013015925884247
+    },
+    {
+      "global_step": 740,
+      "optimizer_step": 92,
+      "lr": 4.7130130375625305e-05,
+      "lm_loss": 0.15300190448760986,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.15300190448760986
+    },
+    {
+      "global_step": 760,
+      "optimizer_step": 95,
+      "lr": 4.6880585547718845e-05,
+      "lm_loss": 0.3297501802444458,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.3297501802444458
+    },
+    {
+      "global_step": 780,
+      "optimizer_step": 97,
+      "lr": 4.670883565255264e-05,
+      "lm_loss": 0.017043394967913628,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.017043394967913628
+    },
+    {
+      "global_step": 800,
+      "optimizer_step": 100,
+      "lr": 4.6443215250006806e-05,
+      "lm_loss": 0.3523687720298767,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.3523687720298767
+    },
+    {
+      "global_step": 820,
+      "optimizer_step": 102,
+      "lr": 4.626085540370762e-05,
+      "lm_loss": 0.012201391160488129,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.012201391160488129
+    },
+    {
+      "global_step": 840,
+      "optimizer_step": 105,
+      "lr": 4.597948593966256e-05,
+      "lm_loss": 0.07073015719652176,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.07073015719652176
+    },
+    {
+      "global_step": 860,
+      "optimizer_step": 107,
+      "lr": 4.5786740307563636e-05,
+      "lm_loss": 0.0668209120631218,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0668209120631218
+    },
+    {
+      "global_step": 880,
+      "optimizer_step": 110,
+      "lr": 4.5489967654649474e-05,
+      "lm_loss": 0.1039772778749466,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1039772778749466
+    },
+    {
+      "global_step": 900,
+      "optimizer_step": 112,
+      "lr": 4.528707316878232e-05,
+      "lm_loss": 0.1377195566892624,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1377195566892624
+    },
+    {
+      "global_step": 920,
+      "optimizer_step": 115,
+      "lr": 4.497526213395623e-05,
+      "lm_loss": 0.1522826999425888,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1522826999425888
+    },
+    {
+      "global_step": 940,
+      "optimizer_step": 117,
+      "lr": 4.476246820180259e-05,
+      "lm_loss": 0.6368707418441772,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.6368707418441772
+    },
+    {
+      "global_step": 960,
+      "optimizer_step": 120,
+      "lr": 4.443600207791081e-05,
+      "lm_loss": 0.2829626798629761,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.2829626798629761
+    },
+    {
+      "global_step": 980,
+      "optimizer_step": 122,
+      "lr": 4.421357027581921e-05,
+      "lm_loss": 0.8739344477653503,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.8739344477653503
+    },
+    {
+      "global_step": 1000,
+      "optimizer_step": 125,
+      "lr": 4.38728503704354e-05,
+      "lm_loss": 0.07859532535076141,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.07859532535076141
+    },
+    {
+      "global_step": 1020,
+      "optimizer_step": 127,
+      "lr": 4.364105412207914e-05,
+      "lm_loss": 0.25118908286094666,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.25118908286094666
+    },
+    {
+      "global_step": 1040,
+      "optimizer_step": 130,
+      "lr": 4.328649926419819e-05,
+      "lm_loss": 0.06467773020267487,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.06467773020267487
+    },
+    {
+      "global_step": 1060,
+      "optimizer_step": 132,
+      "lr": 4.304562350447008e-05,
+      "lm_loss": 0.18095380067825317,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.18095380067825317
+    },
+    {
+      "global_step": 1080,
+      "optimizer_step": 135,
+      "lr": 4.267766952966369e-05,
+      "lm_loss": 0.02313944138586521,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.02313944138586521
+    },
+    {
+      "global_step": 1100,
+      "optimizer_step": 137,
+      "lr": 4.242801035442058e-05,
+      "lm_loss": 0.5191475749015808,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.5191475749015808
+    },
+    {
+      "global_step": 1120,
+      "optimizer_step": 140,
+      "lr": 4.204710956908763e-05,
+      "lm_loss": 0.81064772605896,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.81064772605896
+    },
+    {
+      "global_step": 1140,
+      "optimizer_step": 142,
+      "lr": 4.178897387117546e-05,
+      "lm_loss": 0.0647754892706871,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0647754892706871
+    },
+    {
+      "global_step": 1160,
+      "optimizer_step": 145,
+      "lr": 4.139559449654561e-05,
+      "lm_loss": 0.048019688576459885,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.048019688576459885
+    },
+    {
+      "global_step": 1180,
+      "optimizer_step": 147,
+      "lr": 4.1129299588552193e-05,
+      "lm_loss": 0.160772904753685,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.160772904753685
+    },
+    {
+      "global_step": 1200,
+      "optimizer_step": 150,
+      "lr": 4.072392518512623e-05,
+      "lm_loss": 0.4824674129486084,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.4824674129486084
+    },
+    {
+      "global_step": 1220,
+      "optimizer_step": 152,
+      "lr": 4.0449798409325665e-05,
+      "lm_loss": 0.05526652932167053,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.05526652932167053
+    },
+    {
+      "global_step": 1240,
+      "optimizer_step": 155,
+      "lr": 4.0032927282460146e-05,
+      "lm_loss": 0.33561477065086365,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.33561477065086365
+    },
+    {
+      "global_step": 1260,
+      "optimizer_step": 157,
+      "lr": 3.9751305608428205e-05,
+      "lm_loss": 0.06317351758480072,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.06317351758480072
+    },
+    {
+      "global_step": 1280,
+      "optimizer_step": 160,
+      "lr": 3.932345019579495e-05,
+      "lm_loss": 0.07736121863126755,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.07736121863126755
+    },
+    {
+      "global_step": 1300,
+      "optimizer_step": 162,
+      "lr": 3.90346798061902e-05,
+      "lm_loss": 0.13744595646858215,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.13744595646858215
+    },
+    {
+      "global_step": 1320,
+      "optimizer_step": 165,
+      "lr": 3.859636604786372e-05,
+      "lm_loss": 0.18504707515239716,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.18504707515239716
+    },
+    {
+      "global_step": 1340,
+      "optimizer_step": 167,
+      "lr": 3.830080191288342e-05,
+      "lm_loss": 0.20438233017921448,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.20438233017921448
+    },
+    {
+      "global_step": 1360,
+      "optimizer_step": 170,
+      "lr": 3.785256860483054e-05,
+      "lm_loss": 0.06403923779726028,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.06403923779726028
+    },
+    {
+      "global_step": 1380,
+      "optimizer_step": 172,
+      "lr": 3.755057404586455e-05,
+      "lm_loss": 0.20535588264465332,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.20535588264465332
+    },
+    {
+      "global_step": 1400,
+      "optimizer_step": 175,
+      "lr": 3.7092972177631e-05,
+      "lm_loss": 0.14258944988250732,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.14258944988250732
+    },
+    {
+      "global_step": 1420,
+      "optimizer_step": 177,
+      "lr": 3.678491842064995e-05,
+      "lm_loss": 0.10490252077579498,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.10490252077579498
+    },
+    {
+      "global_step": 1440,
+      "optimizer_step": 180,
+      "lr": 3.631851049805793e-05,
+      "lm_loss": 0.0849061980843544,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0849061980843544
+    },
+    {
+      "global_step": 1460,
+      "optimizer_step": 182,
+      "lr": 3.600477621728473e-05,
+      "lm_loss": 0.016845015808939934,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.016845015808939934
+    },
+    {
+      "global_step": 1480,
+      "optimizer_step": 185,
+      "lr": 3.5530135570974274e-05,
+      "lm_loss": 0.03456898778676987,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.03456898778676987
+    },
+    {
+      "global_step": 1500,
+      "optimizer_step": 187,
+      "lr": 3.521110642339991e-05,
+      "lm_loss": 0.18856661021709442,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.18856661021709442
+    },
+    {
+      "global_step": 1520,
+      "optimizer_step": 190,
+      "lr": 3.4728816504063726e-05,
+      "lm_loss": 0.15951912105083466,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.15951912105083466
+    },
+    {
+      "global_step": 1540,
+      "optimizer_step": 192,
+      "lr": 3.440488465537943e-05,
+      "lm_loss": 0.02313467487692833,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.02313467487692833
+    },
+    {
+      "global_step": 1560,
+      "optimizer_step": 195,
+      "lr": 3.391553831655782e-05,
+      "lm_loss": 0.15202517807483673,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.15202517807483673
+    },
+    {
+      "global_step": 1580,
+      "optimizer_step": 197,
+      "lr": 3.358710195908653e-05,
+      "lm_loss": 0.12373112142086029,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.12373112142086029
+    },
+    {
+      "global_step": 1600,
+      "optimizer_step": 200,
+      "lr": 3.309130072840379e-05,
+      "lm_loss": 0.36436015367507935,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.36436015367507935
+    },
+    {
+      "global_step": 1620,
+      "optimizer_step": 202,
+      "lr": 3.2758763591623295e-05,
+      "lm_loss": 0.05805564671754837,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.05805564671754837
+    },
+    {
+      "global_step": 1640,
+      "optimizer_step": 205,
+      "lr": 3.225711693136156e-05,
+      "lm_loss": 0.303356409072876,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.303356409072876
+    },
+    {
+      "global_step": 1660,
+      "optimizer_step": 207,
+      "lr": 3.1920887785621235e-05,
+      "lm_loss": 0.010578176937997341,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.010578176937997341
+    },
+    {
+      "global_step": 1680,
+      "optimizer_step": 210,
+      "lr": 3.141401234354067e-05,
+      "lm_loss": 0.4240923225879669,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.4240923225879669
+    },
+    {
+      "global_step": 1700,
+      "optimizer_step": 212,
+      "lr": 3.10745044975816e-05,
+      "lm_loss": 0.02695256471633911,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.02695256471633911
+    },
+    {
+      "global_step": 1720,
+      "optimizer_step": 215,
+      "lr": 3.056302334890786e-05,
+      "lm_loss": 0.5630366206169128,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.5630366206169128
+    },
+    {
+      "global_step": 1740,
+      "optimizer_step": 217,
+      "lr": 3.022065414180425e-05,
+      "lm_loss": 0.04303278028964996,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.04303278028964996
+    },
+    {
+      "global_step": 1760,
+      "optimizer_step": 220,
+      "lr": 2.9705196023315006e-05,
+      "lm_loss": 0.31514525413513184,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.31514525413513184
+    },
+    {
+      "global_step": 1780,
+      "optimizer_step": 222,
+      "lr": 2.9360386311461275e-05,
+      "lm_loss": 0.10446590185165405,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.10446590185165405
+    },
+    {
+      "global_step": 1800,
+      "optimizer_step": 225,
+      "lr": 2.884158484861325e-05,
+      "lm_loss": 0.06876020133495331,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.06876020133495331
+    },
+    {
+      "global_step": 1820,
+      "optimizer_step": 227,
+      "lr": 2.849475848838749e-05,
+      "lm_loss": 0.03532234951853752,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.03532234951853752
+    },
+    {
+      "global_step": 1840,
+      "optimizer_step": 230,
+      "lr": 2.7973251416434175e-05,
+      "lm_loss": 0.2705795168876648,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.2705795168876648
+    },
+    {
+      "global_step": 1860,
+      "optimizer_step": 232,
+      "lr": 2.7624834743173867e-05,
+      "lm_loss": 0.39659398794174194,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.39659398794174194
+    },
+    {
+      "global_step": 1880,
+      "optimizer_step": 235,
+      "lr": 2.710126312323119e-05,
+      "lm_loss": 0.23329633474349976,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.23329633474349976
+    },
+    {
+      "global_step": 1900,
+      "optimizer_step": 237,
+      "lr": 2.6751684427161683e-05,
+      "lm_loss": 0.012159268371760845,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.012159268371760845
+    },
+    {
+      "global_step": 1920,
+      "optimizer_step": 240,
+      "lr": 2.6226691858185454e-05,
+      "lm_loss": 0.005995640531182289,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.005995640531182289
+    },
+    {
+      "global_step": 1940,
+      "optimizer_step": 242,
+      "lr": 2.587638085794541e-05,
+      "lm_loss": 0.17683321237564087,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.17683321237564087
+    },
+    {
+      "global_step": 1960,
+      "optimizer_step": 245,
+      "lr": 2.535061268558906e-05,
+      "lm_loss": 0.1525726616382599,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1525726616382599
+    },
+    {
+      "global_step": 1980,
+      "optimizer_step": 247,
+      "lr": 2.5e-05,
+      "lm_loss": 0.01789039745926857,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.01789039745926857
+    },
+    {
+      "global_step": 2000,
+      "optimizer_step": 250,
+      "lr": 2.447410252332528e-05,
+      "lm_loss": 0.08553674817085266,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.08553674817085266
+    },
+    {
+      "global_step": 2020,
+      "optimizer_step": 252,
+      "lr": 2.4123619142054597e-05,
+      "lm_loss": 0.055654481053352356,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.055654481053352356
+    },
+    {
+      "global_step": 2040,
+      "optimizer_step": 255,
+      "lr": 2.3598238819070202e-05,
+      "lm_loss": 0.10190901905298233,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.10190901905298233
+    },
+    {
+      "global_step": 2060,
+      "optimizer_step": 257,
+      "lr": 2.3248315572838316e-05,
+      "lm_loss": 0.02517201378941536,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.02517201378941536
+    },
+    {
+      "global_step": 2080,
+      "optimizer_step": 260,
+      "lr": 2.272409822584325e-05,
+      "lm_loss": 0.16173063218593597,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.16173063218593597
+    },
+    {
+      "global_step": 2100,
+      "optimizer_step": 262,
+      "lr": 2.237516525682614e-05,
+      "lm_loss": 0.08013267815113068,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.08013267815113068
+    },
+    {
+      "global_step": 2120,
+      "optimizer_step": 265,
+      "lr": 2.185275527853437e-05,
+      "lm_loss": 0.004093450028449297,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.004093450028449297
+    },
+    {
+      "global_step": 2140,
+      "optimizer_step": 267,
+      "lr": 2.1505241511612522e-05,
+      "lm_loss": 0.166304811835289,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.166304811835289
+    },
+    {
+      "global_step": 2160,
+      "optimizer_step": 270,
+      "lr": 2.0985281073035076e-05,
+      "lm_loss": 0.6515223979949951,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.6515223979949951
+    },
+    {
+      "global_step": 2180,
+      "optimizer_step": 272,
+      "lr": 2.0639613688538734e-05,
+      "lm_loss": 0.1393578201532364,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1393578201532364
+    },
+    {
+      "global_step": 2200,
+      "optimizer_step": 275,
+      "lr": 2.0122741949596797e-05,
+      "lm_loss": 0.003889517392963171,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.003889517392963171
+    },
+    {
+      "global_step": 2220,
+      "optimizer_step": 277,
+      "lr": 1.977934585819576e-05,
+      "lm_loss": 0.019987229257822037,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.019987229257822037
+    },
+    {
+      "global_step": 2240,
+      "optimizer_step": 280,
+      "lr": 1.9266198182035058e-05,
+      "lm_loss": 0.4692233204841614,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.4692233204841614
+    },
+    {
+      "global_step": 2260,
+      "optimizer_step": 282,
+      "lr": 1.8925495502418406e-05,
+      "lm_loss": 0.010459158569574356,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.010459158569574356
+    },
+    {
+      "global_step": 2280,
+      "optimizer_step": 285,
+      "lr": 1.841670267439088e-05,
+      "lm_loss": 0.20777301490306854,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.20777301490306854
+    },
+    {
+      "global_step": 2300,
+      "optimizer_step": 287,
+      "lr": 1.8079112214378768e-05,
+      "lm_loss": 0.011992229148745537,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.011992229148745537
+    },
+    {
+      "global_step": 2320,
+      "optimizer_step": 290,
+      "lr": 1.757529966665141e-05,
+      "lm_loss": 0.012514831498265266,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.012514831498265266
+    },
+    {
+      "global_step": 2340,
+      "optimizer_step": 292,
+      "lr": 1.72412364083767e-05,
+      "lm_loss": 0.0024862894788384438,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0024862894788384438
+    },
+    {
+      "global_step": 2360,
+      "optimizer_step": 295,
+      "lr": 1.6743023451120832e-05,
+      "lm_loss": 0.11175425350666046,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.11175425350666046
+    },
+    {
+      "global_step": 2380,
+      "optimizer_step": 297,
+      "lr": 1.641289804091347e-05,
+      "lm_loss": 0.014667794108390808,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.014667794108390808
+    },
+    {
+      "global_step": 2400,
+      "optimizer_step": 300,
+      "lr": 1.5920897101019383e-05,
+      "lm_loss": 0.036246757954359055,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.036246757954359055
+    },
+    {
+      "global_step": 2420,
+      "optimizer_step": 302,
+      "lr": 1.5595115344620576e-05,
+      "lm_loss": 0.1119471862912178,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1119471862912178
+    },
+    {
+      "global_step": 2440,
+      "optimizer_step": 305,
+      "lr": 1.5109931212873534e-05,
+      "lm_loss": 0.31454402208328247,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.31454402208328247
+    },
+    {
+      "global_step": 2460,
+      "optimizer_step": 307,
+      "lr": 1.4788893576600099e-05,
+      "lm_loss": 0.024080762639641762,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.024080762639641762
+    },
+    {
+      "global_step": 2480,
+      "optimizer_step": 310,
+      "lr": 1.4311122664242954e-05,
+      "lm_loss": 0.010775794275105,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.010775794275105
+    },
+    {
+      "global_step": 2500,
+      "optimizer_step": 312,
+      "lr": 1.399522378271527e-05,
+      "lm_loss": 0.06494537740945816,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.06494537740945816
+    },
+    {
+      "global_step": 2520,
+      "optimizer_step": 315,
+      "lr": 1.3525453388311554e-05,
+      "lm_loss": 0.003908059559762478,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.003908059559762478
+    },
+    {
+      "global_step": 2540,
+      "optimizer_step": 317,
+      "lr": 1.3215081579350058e-05,
+      "lm_loss": 0.0271071195602417,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0271071195602417
+    },
+    {
+      "global_step": 2560,
+      "optimizer_step": 320,
+      "lr": 1.2753889166848909e-05,
+      "lm_loss": 0.10449909418821335,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.10449909418821335
+    },
+    {
+      "global_step": 2580,
+      "optimizer_step": 322,
+      "lr": 1.2449425954135452e-05,
+      "lm_loss": 0.002020007697865367,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.002020007697865367
+    },
+    {
+      "global_step": 2600,
+      "optimizer_step": 325,
+      "lr": 1.1997378443025634e-05,
+      "lm_loss": 0.00576969888061285,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.00576969888061285
+    },
+    {
+      "global_step": 2620,
+      "optimizer_step": 327,
+      "lr": 1.1699198087116589e-05,
+      "lm_loss": 0.014804661273956299,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.014804661273956299
+    },
+    {
+      "global_step": 2640,
+      "optimizer_step": 330,
+      "lr": 1.1256851155542298e-05,
+      "lm_loss": 0.0035301451571285725,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0035301451571285725
+    },
+    {
+      "global_step": 2660,
+      "optimizer_step": 332,
+      "lr": 1.0965320193809808e-05,
+      "lm_loss": 0.12298412621021271,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.12298412621021271
+    },
+    {
+      "global_step": 2680,
+      "optimizer_step": 335,
+      "lr": 1.0533217595504858e-05,
+      "lm_loss": 0.1802627593278885,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1802627593278885
+    },
+    {
+      "global_step": 2700,
+      "optimizer_step": 337,
+      "lr": 1.0248694391571801e-05,
+      "lm_loss": 0.016198735684156418,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.016198735684156418
+    },
+    {
+      "global_step": 2720,
+      "optimizer_step": 340,
+      "lr": 9.82736728745186e-06,
+      "lm_loss": 0.11246120929718018,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.11246120929718018
+    },
+    {
+      "global_step": 2740,
+      "optimizer_step": 342,
+      "lr": 9.550201590674343e-06,
+      "lm_loss": 0.15225285291671753,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.15225285291671753
+    },
+    {
+      "global_step": 2760,
+      "optimizer_step": 345,
+      "lr": 9.140167895908867e-06,
+      "lm_loss": 0.1835518181324005,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.1835518181324005
+    },
+    {
+      "global_step": 2780,
+      "optimizer_step": 347,
+      "lr": 8.870700411447816e-06,
+      "lm_loss": 0.20481030642986298,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.20481030642986298
+    },
+    {
+      "global_step": 2800,
+      "optimizer_step": 350,
+      "lr": 8.472464158814255e-06,
+      "lm_loss": 0.21467795968055725,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.21467795968055725
+    },
+    {
+      "global_step": 2820,
+      "optimizer_step": 352,
+      "lr": 8.211026128824539e-06,
+      "lm_loss": 0.2778780162334442,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.2778780162334442
+    },
+    {
+      "global_step": 2840,
+      "optimizer_step": 355,
+      "lr": 7.825076849127458e-06,
+      "lm_loss": 0.007937369868159294,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.007937369868159294
+    },
+    {
+      "global_step": 2860,
+      "optimizer_step": 357,
+      "lr": 7.571989645579419e-06,
+      "lm_loss": 0.23060961067676544,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.23060961067676544
+    },
+    {
+      "global_step": 2880,
+      "optimizer_step": 360,
+      "lr": 7.198801765896115e-06,
+      "lm_loss": 0.0005977630498819053,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0005977630498819053
+    },
+    {
+      "global_step": 2900,
+      "optimizer_step": 362,
+      "lr": 6.9543764955299265e-06,
+      "lm_loss": 0.008497863076627254,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.008497863076627254
+    },
+    {
+      "global_step": 2920,
+      "optimizer_step": 365,
+      "lr": 6.594408756022272e-06,
+      "lm_loss": 0.009758133441209793,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.009758133441209793
+    },
+    {
+      "global_step": 2940,
+      "optimizer_step": 367,
+      "lr": 6.358945877920861e-06,
+      "lm_loss": 0.044932521879673004,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.044932521879673004
+    },
+    {
+      "global_step": 2960,
+      "optimizer_step": 370,
+      "lr": 6.012640767930217e-06,
+      "lm_loss": 0.00260924082249403,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.00260924082249403
+    },
+    {
+      "global_step": 2980,
+      "optimizer_step": 372,
+      "lr": 5.7864297241808e-06,
+      "lm_loss": 0.08868487179279327,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.08868487179279327
+    },
+    {
+      "global_step": 3000,
+      "optimizer_step": 375,
+      "lr": 5.454212938299255e-06,
+      "lm_loss": 0.0025052388664335012,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0025052388664335012
+    },
+    {
+      "global_step": 3020,
+      "optimizer_step": 377,
+      "lr": 5.237531798197415e-06,
+      "lm_loss": 0.0014659692533314228,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0014659692533314228
+    },
+    {
+      "global_step": 3040,
+      "optimizer_step": 380,
+      "lr": 4.91981171298388e-06,
+      "lm_loss": 0.03283361345529556,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.03283361345529556
+    },
+    {
+      "global_step": 3060,
+      "optimizer_step": 382,
+      "lr": 4.712926831217685e-06,
+      "lm_loss": 0.13796555995941162,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.13796555995941162
+    },
+    {
+      "global_step": 3080,
+      "optimizer_step": 385,
+      "lr": 4.410094003202134e-06,
+      "lm_loss": 0.18190844357013702,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.18190844357013702
+    },
+    {
+      "global_step": 3100,
+      "optimizer_step": 387,
+      "lr": 4.213259692436367e-06,
+      "lm_loss": 0.0055811768397688866,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0055811768397688866
+    },
+    {
+      "global_step": 3120,
+      "optimizer_step": 390,
+      "lr": 3.925686378029331e-06,
+      "lm_loss": 0.12543368339538574,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.12543368339538574
+    },
+    {
+      "global_step": 3140,
+      "optimizer_step": 392,
+      "lr": 3.739144596292382e-06,
+      "lm_loss": 0.004815898835659027,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.004815898835659027
+    },
+    {
+      "global_step": 3160,
+      "optimizer_step": 395,
+      "lr": 3.4671842941897765e-06,
+      "lm_loss": 0.028516722843050957,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.028516722843050957
+    },
+    {
+      "global_step": 3180,
+      "optimizer_step": 397,
+      "lr": 3.2911643474473646e-06,
+      "lm_loss": 0.10872653871774673,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.10872653871774673
+    },
+    {
+      "global_step": 3200,
+      "optimizer_step": 400,
+      "lr": 3.035151364093214e-06,
+      "lm_loss": 0.003767058253288269,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.003767058253288269
+    },
+    {
+      "global_step": 3220,
+      "optimizer_step": 402,
+      "lr": 2.869869624374699e-06,
+      "lm_loss": 0.0024846738670021296,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0024846738670021296
+    },
+    {
+      "global_step": 3240,
+      "optimizer_step": 405,
+      "lr": 2.6301186630158485e-06,
+      "lm_loss": 0.0013623477425426245,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0013623477425426245
+    },
+    {
+      "global_step": 3260,
+      "optimizer_step": 407,
+      "lr": 2.475778302439524e-06,
+      "lm_loss": 0.006008035968989134,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.006008035968989134
+    },
+    {
+      "global_step": 3280,
+      "optimizer_step": 410,
+      "lr": 2.2525840762775863e-06,
+      "lm_loss": 0.0055444510653615,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0055444510653615
+    },
+    {
+      "global_step": 3300,
+      "optimizer_step": 412,
+      "lr": 2.1093748173018848e-06,
+      "lm_loss": 0.0006941308383829892,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0006941308383829892
+    },
+    {
+      "global_step": 3320,
+      "optimizer_step": 415,
+      "lr": 1.9030116872178316e-06,
+      "lm_loss": 0.04510759189724922,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.04510759189724922
+    },
+    {
+      "global_step": 3340,
+      "optimizer_step": 417,
+      "lr": 1.771109569425547e-06,
+      "lm_loss": 0.4448617994785309,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.4448617994785309
+    },
+    {
+      "global_step": 3360,
+      "optimizer_step": 420,
+      "lr": 1.5818312067223672e-06,
+      "lm_loss": 0.0230400487780571,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0230400487780571
+    },
+    {
+      "global_step": 3380,
+      "optimizer_step": 422,
+      "lr": 1.4613983704244826e-06,
+      "lm_loss": 0.04516194760799408,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.04516194760799408
+    },
+    {
+      "global_step": 3400,
+      "optimizer_step": 425,
+      "lr": 1.2894374450024338e-06,
+      "lm_loss": 0.004301550332456827,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.004301550332456827
+    },
+    {
+      "global_step": 3420,
+      "optimizer_step": 427,
+      "lr": 1.180621931927592e-06,
+      "lm_loss": 0.0042705656960606575,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0042705656960606575
+    },
+    {
+      "global_step": 3440,
+      "optimizer_step": 430,
+      "lr": 1.0261898262753812e-06,
+      "lm_loss": 0.00825844332575798,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.00825844332575798
+    },
+    {
+      "global_step": 3460,
+      "optimizer_step": 432,
+      "lr": 9.291253975900139e-07,
+      "lm_loss": 0.0010993486503139138,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0010993486503139138
+    },
+    {
+      "global_step": 3480,
+      "optimizer_step": 435,
+      "lr": 7.924119469434665e-07,
+      "lm_loss": 0.0028111122082918882,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0028111122082918882
+    },
+    {
+      "global_step": 3500,
+      "optimizer_step": 437,
+      "lr": 7.072179188262251e-07,
+      "lm_loss": 0.0014011103194206953,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0014011103194206953
+    },
+    {
+      "global_step": 3520,
+      "optimizer_step": 440,
+      "lr": 5.883911778138324e-07,
+      "lm_loss": 0.20520374178886414,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.20520374178886414
+    },
+    {
+      "global_step": 3540,
+      "optimizer_step": 442,
+      "lr": 5.151722747865434e-07,
+      "lm_loss": 0.07059439271688461,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.07059439271688461
+    },
+    {
+      "global_step": 3560,
+      "optimizer_step": 445,
+      "lr": 4.1437831084878974e-07,
+      "lm_loss": 0.19064895808696747,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.19064895808696747
+    },
+    {
+      "global_step": 3580,
+      "optimizer_step": 447,
+      "lr": 3.5322453704410286e-07,
+      "lm_loss": 0.0009202686487697065,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0009202686487697065
+    },
+    {
+      "global_step": 3600,
+      "optimizer_step": 450,
+      "lr": 2.705872508804747e-07,
+      "lm_loss": 0.038131456822156906,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.038131456822156906
+    },
+    {
+      "global_step": 3620,
+      "optimizer_step": 452,
+      "lr": 2.215737794045436e-07,
+      "lm_loss": 0.001025449950248003,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.001025449950248003
+    },
+    {
+      "global_step": 3640,
+      "optimizer_step": 455,
+      "lr": 1.571947526689349e-07,
+      "lm_loss": 0.0016507974360138178,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0016507974360138178
+    },
+    {
+      "global_step": 3660,
+      "optimizer_step": 457,
+      "lr": 1.2038183319507955e-07,
+      "lm_loss": 0.01071799173951149,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.01071799173951149
+    },
+    {
+      "global_step": 3680,
+      "optimizer_step": 460,
+      "lr": 7.434020362682704e-08,
+      "lm_loss": 0.0010236427187919617,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0010236427187919617
+    },
+    {
+      "global_step": 3700,
+      "optimizer_step": 462,
+      "lr": 4.977308833379324e-08,
+      "lm_loss": 0.013086177408695221,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.013086177408695221
+    },
+    {
+      "global_step": 3720,
+      "optimizer_step": 465,
+      "lr": 2.2125452477828047e-08,
+      "lm_loss": 0.0018851282075047493,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.0018851282075047493
+    },
+    {
+      "global_step": 3740,
+      "optimizer_step": 467,
+      "lr": 9.834340423678368e-09,
+      "lm_loss": 0.007189369294792414,
+      "order_loss": 0.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 0.007189369294792414
+    }
+  ]
+}

cls_gemma_orderkd_first_l01/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_first_l01/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_first_l01/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_orderkd_first_l01/checkpoint-200/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_first_l01/checkpoint-200/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_first_l01/checkpoint-200/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_orderkd_first_l01/checkpoint-400/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_first_l01/checkpoint-400/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_first_l01/checkpoint-400/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_orderkd_first_l01/train_metrics.json

@@ -0,0 +1,1874 @@
+{
+  "loss_log": [
+    {
+      "global_step": 20,
+      "optimizer_step": 2,
+      "lr": 4.347826086956522e-06,
+      "lm_loss": 3.1184232234954834,
+      "order_loss": 63.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 9.462173461914062
+    },
+    {
+      "global_step": 40,
+      "optimizer_step": 5,
+      "lr": 1.0869565217391305e-05,
+      "lm_loss": 1.0916705131530762,
+      "order_loss": 56.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.747920513153076
+    },
+    {
+      "global_step": 60,
+      "optimizer_step": 7,
+      "lr": 1.5217391304347828e-05,
+      "lm_loss": 0.6041339039802551,
+      "order_loss": 42.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.8228840827941895
+    },
+    {
+      "global_step": 80,
+      "optimizer_step": 10,
+      "lr": 2.173913043478261e-05,
+      "lm_loss": 1.8147910833358765,
+      "order_loss": 55.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.346041202545166
+    },
+    {
+      "global_step": 100,
+      "optimizer_step": 12,
+      "lr": 2.608695652173913e-05,
+      "lm_loss": 0.5634512305259705,
+      "order_loss": 43.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.907201290130615
+    },
+    {
+      "global_step": 120,
+      "optimizer_step": 15,
+      "lr": 3.260869565217392e-05,
+      "lm_loss": 2.721200942993164,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 8.158700942993164
+    },
+    {
+      "global_step": 140,
+      "optimizer_step": 17,
+      "lr": 3.695652173913043e-05,
+      "lm_loss": 0.6781831979751587,
+      "order_loss": 54.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.084433078765869
+    },
+    {
+      "global_step": 160,
+      "optimizer_step": 20,
+      "lr": 4.347826086956522e-05,
+      "lm_loss": 0.41744333505630493,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.07369327545166
+    },
+    {
+      "global_step": 180,
+      "optimizer_step": 22,
+      "lr": 4.782608695652174e-05,
+      "lm_loss": 0.06321373581886292,
+      "order_loss": 54.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.50071382522583
+    },
+    {
+      "global_step": 200,
+      "optimizer_step": 25,
+      "lr": 4.9997541293989384e-05,
+      "lm_loss": 0.02435319498181343,
+      "order_loss": 54.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.46185302734375
+    },
+    {
+      "global_step": 220,
+      "optimizer_step": 27,
+      "lr": 4.999016565957633e-05,
+      "lm_loss": 0.019743939861655235,
+      "order_loss": 61.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.144743919372559
+    },
+    {
+      "global_step": 240,
+      "optimizer_step": 30,
+      "lr": 4.996988640512931e-05,
+      "lm_loss": 0.3459993302822113,
+      "order_loss": 39.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.314749240875244
+    },
+    {
+      "global_step": 260,
+      "optimizer_step": 32,
+      "lr": 4.995022691166621e-05,
+      "lm_loss": 0.0298320259898901,
+      "order_loss": 62.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.248581886291504
+    },
+    {
+      "global_step": 280,
+      "optimizer_step": 35,
+      "lr": 4.991153735294049e-05,
+      "lm_loss": 0.46130573749542236,
+      "order_loss": 60.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.523805618286133
+    },
+    {
+      "global_step": 300,
+      "optimizer_step": 37,
+      "lr": 4.987961816680492e-05,
+      "lm_loss": 0.1003972664475441,
+      "order_loss": 54.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.569147109985352
+    },
+    {
+      "global_step": 320,
+      "optimizer_step": 40,
+      "lr": 4.9822565862832806e-05,
+      "lm_loss": 0.13187472522258759,
+      "order_loss": 40.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.1943745613098145
+    },
+    {
+      "global_step": 340,
+      "optimizer_step": 42,
+      "lr": 4.9778426220595456e-05,
+      "lm_loss": 0.32335221767425537,
+      "order_loss": 50.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.323352336883545
+    },
+    {
+      "global_step": 360,
+      "optimizer_step": 45,
+      "lr": 4.970308130276272e-05,
+      "lm_loss": 0.7211238741874695,
+      "order_loss": 56.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.377373695373535
+    },
+    {
+      "global_step": 380,
+      "optimizer_step": 47,
+      "lr": 4.96467754629559e-05,
+      "lm_loss": 0.03598100692033768,
+      "order_loss": 41.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.223481178283691
+    },
+    {
+      "global_step": 400,
+      "optimizer_step": 50,
+      "lr": 4.955323054879298e-05,
+      "lm_loss": 0.10630018264055252,
+      "order_loss": 47.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.825050354003906
+    },
+    {
+      "global_step": 420,
+      "optimizer_step": 52,
+      "lr": 4.948482772521345e-05,
+      "lm_loss": 0.11145680397748947,
+      "order_loss": 66.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.705206871032715
+    },
+    {
+      "global_step": 440,
+      "optimizer_step": 55,
+      "lr": 4.937319780454559e-05,
+      "lm_loss": 0.18087011575698853,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.618370056152344
+    },
+    {
+      "global_step": 460,
+      "optimizer_step": 57,
+      "lr": 4.929278208117378e-05,
+      "lm_loss": 0.6563955545425415,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.625145435333252
+    },
+    {
+      "global_step": 480,
+      "optimizer_step": 60,
+      "lr": 4.916320437477008e-05,
+      "lm_loss": 0.32431942224502563,
+      "order_loss": 57.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.043069362640381
+    },
+    {
+      "global_step": 500,
+      "optimizer_step": 62,
+      "lr": 4.907087460240999e-05,
+      "lm_loss": 0.14991332590579987,
+      "order_loss": 61.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.274913311004639
+    },
+    {
+      "global_step": 520,
+      "optimizer_step": 65,
+      "lr": 4.892350839330522e-05,
+      "lm_loss": 0.1010974645614624,
+      "order_loss": 50.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.163597583770752
+    },
+    {
+      "global_step": 540,
+      "optimizer_step": 67,
+      "lr": 4.881937806807241e-05,
+      "lm_loss": 0.15496166050434113,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.779961585998535
+    },
+    {
+      "global_step": 560,
+      "optimizer_step": 70,
+      "lr": 4.865440450576874e-05,
+      "lm_loss": 0.026232725009322166,
+      "order_loss": 56.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.713732719421387
+    },
+    {
+      "global_step": 580,
+      "optimizer_step": 72,
+      "lr": 4.853860162957552e-05,
+      "lm_loss": 0.09959332644939423,
+      "order_loss": 57.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.787093162536621
+    },
+    {
+      "global_step": 600,
+      "optimizer_step": 75,
+      "lr": 4.8356223507364996e-05,
+      "lm_loss": 0.01794126257300377,
+      "order_loss": 52.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.299191474914551
+    },
+    {
+      "global_step": 620,
+      "optimizer_step": 77,
+      "lr": 4.822889043057446e-05,
+      "lm_loss": 0.26736268401145935,
+      "order_loss": 49.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.204862594604492
+    },
+    {
+      "global_step": 640,
+      "optimizer_step": 80,
+      "lr": 4.802933193625593e-05,
+      "lm_loss": 0.007746524177491665,
+      "order_loss": 67.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.695246696472168
+    },
+    {
+      "global_step": 660,
+      "optimizer_step": 82,
+      "lr": 4.789062518269812e-05,
+      "lm_loss": 0.03736429661512375,
+      "order_loss": 52.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.318614482879639
+    },
+    {
+      "global_step": 680,
+      "optimizer_step": 85,
+      "lr": 4.767413162299501e-05,
+      "lm_loss": 0.06276468187570572,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.031514644622803
+    },
+    {
+      "global_step": 700,
+      "optimizer_step": 87,
+      "lr": 4.752422169756048e-05,
+      "lm_loss": 0.22046461701393127,
+      "order_loss": 67.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.970464706420898
+    },
+    {
+      "global_step": 720,
+      "optimizer_step": 90,
+      "lr": 4.729105919657818e-05,
+      "lm_loss": 0.4013015925884247,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.151301383972168
+    },
+    {
+      "global_step": 740,
+      "optimizer_step": 92,
+      "lr": 4.7130130375625305e-05,
+      "lm_loss": 0.15300190448760986,
+      "order_loss": 55.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.65300178527832
+    },
+    {
+      "global_step": 760,
+      "optimizer_step": 95,
+      "lr": 4.6880585547718845e-05,
+      "lm_loss": 0.3297501802444458,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.079750061035156
+    },
+    {
+      "global_step": 780,
+      "optimizer_step": 97,
+      "lr": 4.670883565255264e-05,
+      "lm_loss": 0.017043394967913628,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.267043590545654
+    },
+    {
+      "global_step": 800,
+      "optimizer_step": 100,
+      "lr": 4.6443215250006806e-05,
+      "lm_loss": 0.3523687720298767,
+      "order_loss": 56.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9461188316345215
+    },
+    {
+      "global_step": 820,
+      "optimizer_step": 102,
+      "lr": 4.626085540370762e-05,
+      "lm_loss": 0.012201391160488129,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.793451309204102
+    },
+    {
+      "global_step": 840,
+      "optimizer_step": 105,
+      "lr": 4.597948593966256e-05,
+      "lm_loss": 0.07073015719652176,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.039480209350586
+    },
+    {
+      "global_step": 860,
+      "optimizer_step": 107,
+      "lr": 4.5786740307563636e-05,
+      "lm_loss": 0.0668209120631218,
+      "order_loss": 63.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.441821098327637
+    },
+    {
+      "global_step": 880,
+      "optimizer_step": 110,
+      "lr": 4.5489967654649474e-05,
+      "lm_loss": 0.1039772778749466,
+      "order_loss": 63.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.416477203369141
+    },
+    {
+      "global_step": 900,
+      "optimizer_step": 112,
+      "lr": 4.528707316878232e-05,
+      "lm_loss": 0.1377195566892624,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.075219631195068
+    },
+    {
+      "global_step": 920,
+      "optimizer_step": 115,
+      "lr": 4.497526213395623e-05,
+      "lm_loss": 0.1522826999425888,
+      "order_loss": 55.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.71478271484375
+    },
+    {
+      "global_step": 940,
+      "optimizer_step": 117,
+      "lr": 4.476246820180259e-05,
+      "lm_loss": 0.6368707418441772,
+      "order_loss": 52.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.855620861053467
+    },
+    {
+      "global_step": 960,
+      "optimizer_step": 120,
+      "lr": 4.443600207791081e-05,
+      "lm_loss": 0.2829626798629761,
+      "order_loss": 61.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.439212799072266
+    },
+    {
+      "global_step": 980,
+      "optimizer_step": 122,
+      "lr": 4.421357027581921e-05,
+      "lm_loss": 0.8739344477653503,
+      "order_loss": 55.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.405184268951416
+    },
+    {
+      "global_step": 1000,
+      "optimizer_step": 125,
+      "lr": 4.38728503704354e-05,
+      "lm_loss": 0.07859532535076141,
+      "order_loss": 55.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.641095161437988
+    },
+    {
+      "global_step": 1020,
+      "optimizer_step": 127,
+      "lr": 4.364105412207914e-05,
+      "lm_loss": 0.25118908286094666,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.751189231872559
+    },
+    {
+      "global_step": 1040,
+      "optimizer_step": 130,
+      "lr": 4.328649926419819e-05,
+      "lm_loss": 0.06467773020267487,
+      "order_loss": 52.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.283427715301514
+    },
+    {
+      "global_step": 1060,
+      "optimizer_step": 132,
+      "lr": 4.304562350447008e-05,
+      "lm_loss": 0.18095380067825317,
+      "order_loss": 56.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.7747039794921875
+    },
+    {
+      "global_step": 1080,
+      "optimizer_step": 135,
+      "lr": 4.267766952966369e-05,
+      "lm_loss": 0.02313944138586521,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.273139476776123
+    },
+    {
+      "global_step": 1100,
+      "optimizer_step": 137,
+      "lr": 4.242801035442058e-05,
+      "lm_loss": 0.5191475749015808,
+      "order_loss": 62.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.7066473960876465
+    },
+    {
+      "global_step": 1120,
+      "optimizer_step": 140,
+      "lr": 4.204710956908763e-05,
+      "lm_loss": 0.81064772605896,
+      "order_loss": 62.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.998147964477539
+    },
+    {
+      "global_step": 1140,
+      "optimizer_step": 142,
+      "lr": 4.178897387117546e-05,
+      "lm_loss": 0.0647754892706871,
+      "order_loss": 61.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.158525466918945
+    },
+    {
+      "global_step": 1160,
+      "optimizer_step": 145,
+      "lr": 4.139559449654561e-05,
+      "lm_loss": 0.048019688576459885,
+      "order_loss": 53.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.391769886016846
+    },
+    {
+      "global_step": 1180,
+      "optimizer_step": 147,
+      "lr": 4.1129299588552193e-05,
+      "lm_loss": 0.160772904753685,
+      "order_loss": 61.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.285772800445557
+    },
+    {
+      "global_step": 1200,
+      "optimizer_step": 150,
+      "lr": 4.072392518512623e-05,
+      "lm_loss": 0.4824674129486084,
+      "order_loss": 63.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.8262176513671875
+    },
+    {
+      "global_step": 1220,
+      "optimizer_step": 152,
+      "lr": 4.0449798409325665e-05,
+      "lm_loss": 0.05526652932167053,
+      "order_loss": 52.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.305266380310059
+    },
+    {
+      "global_step": 1240,
+      "optimizer_step": 155,
+      "lr": 4.0032927282460146e-05,
+      "lm_loss": 0.33561477065086365,
+      "order_loss": 60.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.3668646812438965
+    },
+    {
+      "global_step": 1260,
+      "optimizer_step": 157,
+      "lr": 3.9751305608428205e-05,
+      "lm_loss": 0.17131206393241882,
+      "order_loss": 57.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.858811855316162
+    },
+    {
+      "global_step": 1280,
+      "optimizer_step": 160,
+      "lr": 3.932345019579495e-05,
+      "lm_loss": 0.04662935063242912,
+      "order_loss": 61.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.234129428863525
+    },
+    {
+      "global_step": 1300,
+      "optimizer_step": 162,
+      "lr": 3.90346798061902e-05,
+      "lm_loss": 0.050219446420669556,
+      "order_loss": 50.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.081469535827637
+    },
+    {
+      "global_step": 1320,
+      "optimizer_step": 165,
+      "lr": 3.859636604786372e-05,
+      "lm_loss": 0.34347474575042725,
+      "order_loss": 48.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.155974864959717
+    },
+    {
+      "global_step": 1340,
+      "optimizer_step": 167,
+      "lr": 3.830080191288342e-05,
+      "lm_loss": 0.4303498864173889,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.867849826812744
+    },
+    {
+      "global_step": 1360,
+      "optimizer_step": 170,
+      "lr": 3.785256860483054e-05,
+      "lm_loss": 0.07431875914335251,
+      "order_loss": 53.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.418068885803223
+    },
+    {
+      "global_step": 1380,
+      "optimizer_step": 172,
+      "lr": 3.755057404586455e-05,
+      "lm_loss": 0.14596708118915558,
+      "order_loss": 39.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.06784200668335
+    },
+    {
+      "global_step": 1400,
+      "optimizer_step": 175,
+      "lr": 3.7092972177631e-05,
+      "lm_loss": 0.13190682232379913,
+      "order_loss": 65.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.631906986236572
+    },
+    {
+      "global_step": 1420,
+      "optimizer_step": 177,
+      "lr": 3.678491842064995e-05,
+      "lm_loss": 0.15802578628063202,
+      "order_loss": 51.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.283025741577148
+    },
+    {
+      "global_step": 1440,
+      "optimizer_step": 180,
+      "lr": 3.631851049805793e-05,
+      "lm_loss": 0.30685389041900635,
+      "order_loss": 45.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.869353771209717
+    },
+    {
+      "global_step": 1460,
+      "optimizer_step": 182,
+      "lr": 3.600477621728473e-05,
+      "lm_loss": 0.01724999025464058,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.642250061035156
+    },
+    {
+      "global_step": 1480,
+      "optimizer_step": 185,
+      "lr": 3.5530135570974274e-05,
+      "lm_loss": 0.25045639276504517,
+      "order_loss": 50.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.3129563331604
+    },
+    {
+      "global_step": 1500,
+      "optimizer_step": 187,
+      "lr": 3.521110642339991e-05,
+      "lm_loss": 0.4265330731868744,
+      "order_loss": 52.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.707783222198486
+    },
+    {
+      "global_step": 1520,
+      "optimizer_step": 190,
+      "lr": 3.4728816504063726e-05,
+      "lm_loss": 0.23969930410385132,
+      "order_loss": 61.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.427199363708496
+    },
+    {
+      "global_step": 1540,
+      "optimizer_step": 192,
+      "lr": 3.440488465537943e-05,
+      "lm_loss": 0.006110867951065302,
+      "order_loss": 65.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.506110668182373
+    },
+    {
+      "global_step": 1560,
+      "optimizer_step": 195,
+      "lr": 3.391553831655782e-05,
+      "lm_loss": 0.01415237132459879,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.951652526855469
+    },
+    {
+      "global_step": 1580,
+      "optimizer_step": 197,
+      "lr": 3.358710195908653e-05,
+      "lm_loss": 0.2871032953262329,
+      "order_loss": 70.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.287103176116943
+    },
+    {
+      "global_step": 1600,
+      "optimizer_step": 200,
+      "lr": 3.309130072840379e-05,
+      "lm_loss": 0.23844397068023682,
+      "order_loss": 58.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.050943851470947
+    },
+    {
+      "global_step": 1620,
+      "optimizer_step": 202,
+      "lr": 3.2758763591623295e-05,
+      "lm_loss": 0.1546662300825119,
+      "order_loss": 50.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.185916423797607
+    },
+    {
+      "global_step": 1640,
+      "optimizer_step": 205,
+      "lr": 3.225711693136156e-05,
+      "lm_loss": 0.08392491936683655,
+      "order_loss": 60.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.146424770355225
+    },
+    {
+      "global_step": 1660,
+      "optimizer_step": 207,
+      "lr": 3.1920887785621235e-05,
+      "lm_loss": 0.043332748115062714,
+      "order_loss": 61.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.230832576751709
+    },
+    {
+      "global_step": 1680,
+      "optimizer_step": 210,
+      "lr": 3.141401234354067e-05,
+      "lm_loss": 0.13834291696548462,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.10709285736084
+    },
+    {
+      "global_step": 1700,
+      "optimizer_step": 212,
+      "lr": 3.10745044975816e-05,
+      "lm_loss": 0.2134607881307602,
+      "order_loss": 50.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.275960922241211
+    },
+    {
+      "global_step": 1720,
+      "optimizer_step": 215,
+      "lr": 3.056302334890786e-05,
+      "lm_loss": 0.01857478730380535,
+      "order_loss": 74.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.4248247146606445
+    },
+    {
+      "global_step": 1740,
+      "optimizer_step": 217,
+      "lr": 3.022065414180425e-05,
+      "lm_loss": 0.006793432869017124,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.25679349899292
+    },
+    {
+      "global_step": 1760,
+      "optimizer_step": 220,
+      "lr": 2.9705196023315006e-05,
+      "lm_loss": 0.013562669977545738,
+      "order_loss": 59.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9510626792907715
+    },
+    {
+      "global_step": 1780,
+      "optimizer_step": 222,
+      "lr": 2.9360386311461275e-05,
+      "lm_loss": 0.0029967674054205418,
+      "order_loss": 47.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.721746921539307
+    },
+    {
+      "global_step": 1800,
+      "optimizer_step": 225,
+      "lr": 2.884158484861325e-05,
+      "lm_loss": 0.012671629898250103,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.76267147064209
+    },
+    {
+      "global_step": 1820,
+      "optimizer_step": 227,
+      "lr": 2.849475848838749e-05,
+      "lm_loss": 0.0835554227232933,
+      "order_loss": 39.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.036680221557617
+    },
+    {
+      "global_step": 1840,
+      "optimizer_step": 230,
+      "lr": 2.7973251416434175e-05,
+      "lm_loss": 0.05105651170015335,
+      "order_loss": 29.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 3.0041816234588623
+    },
+    {
+      "global_step": 1860,
+      "optimizer_step": 232,
+      "lr": 2.7624834743173867e-05,
+      "lm_loss": 0.5029534101486206,
+      "order_loss": 75.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 8.00295352935791
+    },
+    {
+      "global_step": 1880,
+      "optimizer_step": 235,
+      "lr": 2.710126312323119e-05,
+      "lm_loss": 0.12120761722326279,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.3712077140808105
+    },
+    {
+      "global_step": 1900,
+      "optimizer_step": 237,
+      "lr": 2.6751684427161683e-05,
+      "lm_loss": 0.17923295497894287,
+      "order_loss": 40.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.210483074188232
+    },
+    {
+      "global_step": 1920,
+      "optimizer_step": 240,
+      "lr": 2.6226691858185454e-05,
+      "lm_loss": 0.004263256676495075,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.504263401031494
+    },
+    {
+      "global_step": 1940,
+      "optimizer_step": 242,
+      "lr": 2.587638085794541e-05,
+      "lm_loss": 0.04619055613875389,
+      "order_loss": 56.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.7024407386779785
+    },
+    {
+      "global_step": 1960,
+      "optimizer_step": 245,
+      "lr": 2.535061268558906e-05,
+      "lm_loss": 0.11890175938606262,
+      "order_loss": 71.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.27515172958374
+    },
+    {
+      "global_step": 1980,
+      "optimizer_step": 247,
+      "lr": 2.5e-05,
+      "lm_loss": 0.25617969036102295,
+      "order_loss": 47.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.9436798095703125
+    },
+    {
+      "global_step": 2000,
+      "optimizer_step": 250,
+      "lr": 2.447410252332528e-05,
+      "lm_loss": 0.013576528988778591,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.451076507568359
+    },
+    {
+      "global_step": 2020,
+      "optimizer_step": 252,
+      "lr": 2.4123619142054597e-05,
+      "lm_loss": 0.04630662128329277,
+      "order_loss": 50.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.046306610107422
+    },
+    {
+      "global_step": 2040,
+      "optimizer_step": 255,
+      "lr": 2.3598238819070202e-05,
+      "lm_loss": 0.14818225800991058,
+      "order_loss": 61.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.304432392120361
+    },
+    {
+      "global_step": 2060,
+      "optimizer_step": 257,
+      "lr": 2.3248315572838316e-05,
+      "lm_loss": 0.0021958579309284687,
+      "order_loss": 44.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.470945835113525
+    },
+    {
+      "global_step": 2080,
+      "optimizer_step": 260,
+      "lr": 2.272409822584325e-05,
+      "lm_loss": 0.002902108011767268,
+      "order_loss": 43.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.377902030944824
+    },
+    {
+      "global_step": 2100,
+      "optimizer_step": 262,
+      "lr": 2.237516525682614e-05,
+      "lm_loss": 0.016654551029205322,
+      "order_loss": 47.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.7041544914245605
+    },
+    {
+      "global_step": 2120,
+      "optimizer_step": 265,
+      "lr": 2.185275527853437e-05,
+      "lm_loss": 0.012752072885632515,
+      "order_loss": 43.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.325252056121826
+    },
+    {
+      "global_step": 2140,
+      "optimizer_step": 267,
+      "lr": 2.1505241511612522e-05,
+      "lm_loss": 0.009639913216233253,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.947139739990234
+    },
+    {
+      "global_step": 2160,
+      "optimizer_step": 270,
+      "lr": 2.0985281073035076e-05,
+      "lm_loss": 0.11398650705814362,
+      "order_loss": 44.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.551486492156982
+    },
+    {
+      "global_step": 2180,
+      "optimizer_step": 272,
+      "lr": 2.0639613688538734e-05,
+      "lm_loss": 0.0007875743904151022,
+      "order_loss": 51.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.094537734985352
+    },
+    {
+      "global_step": 2200,
+      "optimizer_step": 275,
+      "lr": 2.0122741949596797e-05,
+      "lm_loss": 0.009382687509059906,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.759382724761963
+    },
+    {
+      "global_step": 2220,
+      "optimizer_step": 277,
+      "lr": 1.977934585819576e-05,
+      "lm_loss": 0.2916325628757477,
+      "order_loss": 48.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.104132652282715
+    },
+    {
+      "global_step": 2240,
+      "optimizer_step": 280,
+      "lr": 1.9266198182035058e-05,
+      "lm_loss": 0.006090455688536167,
+      "order_loss": 61.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.1623406410217285
+    },
+    {
+      "global_step": 2260,
+      "optimizer_step": 282,
+      "lr": 1.8925495502418406e-05,
+      "lm_loss": 0.06058869510889053,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.685588836669922
+    },
+    {
+      "global_step": 2280,
+      "optimizer_step": 285,
+      "lr": 1.841670267439088e-05,
+      "lm_loss": 0.0836532935500145,
+      "order_loss": 51.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.271153450012207
+    },
+    {
+      "global_step": 2300,
+      "optimizer_step": 287,
+      "lr": 1.8079112214378768e-05,
+      "lm_loss": 0.05897977575659752,
+      "order_loss": 58.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9339799880981445
+    },
+    {
+      "global_step": 2320,
+      "optimizer_step": 290,
+      "lr": 1.757529966665141e-05,
+      "lm_loss": 0.1328563094139099,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.789106369018555
+    },
+    {
+      "global_step": 2340,
+      "optimizer_step": 292,
+      "lr": 1.72412364083767e-05,
+      "lm_loss": 0.044375427067279816,
+      "order_loss": 63.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.356875419616699
+    },
+    {
+      "global_step": 2360,
+      "optimizer_step": 295,
+      "lr": 1.6743023451120832e-05,
+      "lm_loss": 0.18005213141441345,
+      "order_loss": 58.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.992552280426025
+    },
+    {
+      "global_step": 2380,
+      "optimizer_step": 297,
+      "lr": 1.641289804091347e-05,
+      "lm_loss": 0.07155724614858627,
+      "order_loss": 48.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.88405704498291
+    },
+    {
+      "global_step": 2400,
+      "optimizer_step": 300,
+      "lr": 1.5920897101019383e-05,
+      "lm_loss": 0.21017329394817352,
+      "order_loss": 48.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.0226731300354
+    },
+    {
+      "global_step": 2420,
+      "optimizer_step": 302,
+      "lr": 1.5595115344620576e-05,
+      "lm_loss": 0.013050842098891735,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.7630510330200195
+    },
+    {
+      "global_step": 2440,
+      "optimizer_step": 305,
+      "lr": 1.5109931212873534e-05,
+      "lm_loss": 0.022372275590896606,
+      "order_loss": 61.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.116122245788574
+    },
+    {
+      "global_step": 2460,
+      "optimizer_step": 307,
+      "lr": 1.4788893576600099e-05,
+      "lm_loss": 0.2720758318901062,
+      "order_loss": 64.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.709575653076172
+    },
+    {
+      "global_step": 2480,
+      "optimizer_step": 310,
+      "lr": 1.4311122664242954e-05,
+      "lm_loss": 0.0060526239685714245,
+      "order_loss": 51.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.162302494049072
+    },
+    {
+      "global_step": 2500,
+      "optimizer_step": 312,
+      "lr": 1.399522378271527e-05,
+      "lm_loss": 0.8424654603004456,
+      "order_loss": 57.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.529965400695801
+    },
+    {
+      "global_step": 2520,
+      "optimizer_step": 315,
+      "lr": 1.3525453388311554e-05,
+      "lm_loss": 0.12010786682367325,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.057607650756836
+    },
+    {
+      "global_step": 2540,
+      "optimizer_step": 317,
+      "lr": 1.3215081579350058e-05,
+      "lm_loss": 0.6840029954910278,
+      "order_loss": 47.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.371502876281738
+    },
+    {
+      "global_step": 2560,
+      "optimizer_step": 320,
+      "lr": 1.2753889166848909e-05,
+      "lm_loss": 0.005617386195808649,
+      "order_loss": 63.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.318117618560791
+    },
+    {
+      "global_step": 2580,
+      "optimizer_step": 322,
+      "lr": 1.2449425954135452e-05,
+      "lm_loss": 0.009596875868737698,
+      "order_loss": 46.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.603346824645996
+    },
+    {
+      "global_step": 2600,
+      "optimizer_step": 325,
+      "lr": 1.1997378443025634e-05,
+      "lm_loss": 0.006365991663187742,
+      "order_loss": 50.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.068865776062012
+    },
+    {
+      "global_step": 2620,
+      "optimizer_step": 327,
+      "lr": 1.1699198087116589e-05,
+      "lm_loss": 0.0015480580041185021,
+      "order_loss": 47.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.7515482902526855
+    },
+    {
+      "global_step": 2640,
+      "optimizer_step": 330,
+      "lr": 1.1256851155542298e-05,
+      "lm_loss": 0.0021843877620995045,
+      "order_loss": 56.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.6896843910217285
+    },
+    {
+      "global_step": 2660,
+      "optimizer_step": 332,
+      "lr": 1.0965320193809808e-05,
+      "lm_loss": 0.01388256810605526,
+      "order_loss": 52.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.263882637023926
+    },
+    {
+      "global_step": 2680,
+      "optimizer_step": 335,
+      "lr": 1.0533217595504858e-05,
+      "lm_loss": 0.0009903698228299618,
+      "order_loss": 53.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.344740390777588
+    },
+    {
+      "global_step": 2700,
+      "optimizer_step": 337,
+      "lr": 1.0248694391571801e-05,
+      "lm_loss": 0.0581960529088974,
+      "order_loss": 51.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.183196067810059
+    },
+    {
+      "global_step": 2720,
+      "optimizer_step": 340,
+      "lr": 9.82736728745186e-06,
+      "lm_loss": 0.05928004905581474,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.809279918670654
+    },
+    {
+      "global_step": 2740,
+      "optimizer_step": 342,
+      "lr": 9.550201590674343e-06,
+      "lm_loss": 0.00916436966508627,
+      "order_loss": 51.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.165414333343506
+    },
+    {
+      "global_step": 2760,
+      "optimizer_step": 345,
+      "lr": 9.140167895908867e-06,
+      "lm_loss": 0.02078745700418949,
+      "order_loss": 45.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.552037239074707
+    },
+    {
+      "global_step": 2780,
+      "optimizer_step": 347,
+      "lr": 8.870700411447816e-06,
+      "lm_loss": 0.0645354613661766,
+      "order_loss": 51.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.252035617828369
+    },
+    {
+      "global_step": 2800,
+      "optimizer_step": 350,
+      "lr": 8.472464158814255e-06,
+      "lm_loss": 0.03742121905088425,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.537421226501465
+    },
+    {
+      "global_step": 2820,
+      "optimizer_step": 352,
+      "lr": 8.211026128824539e-06,
+      "lm_loss": 0.004177003167569637,
+      "order_loss": 56.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.597927093505859
+    },
+    {
+      "global_step": 2840,
+      "optimizer_step": 355,
+      "lr": 7.825076849127458e-06,
+      "lm_loss": 0.017688076943159103,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.673938274383545
+    },
+    {
+      "global_step": 2860,
+      "optimizer_step": 357,
+      "lr": 7.571989645579419e-06,
+      "lm_loss": 0.0998288244009018,
+      "order_loss": 60.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.099828720092773
+    },
+    {
+      "global_step": 2880,
+      "optimizer_step": 360,
+      "lr": 7.198801765896115e-06,
+      "lm_loss": 0.24300594627857208,
+      "order_loss": 64.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.649255752563477
+    },
+    {
+      "global_step": 2900,
+      "optimizer_step": 362,
+      "lr": 6.9543764955299265e-06,
+      "lm_loss": 0.03150952234864235,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.9690093994140625
+    },
+    {
+      "global_step": 2920,
+      "optimizer_step": 365,
+      "lr": 6.594408756022272e-06,
+      "lm_loss": 0.1402754932641983,
+      "order_loss": 58.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.952775478363037
+    },
+    {
+      "global_step": 2940,
+      "optimizer_step": 367,
+      "lr": 6.358945877920861e-06,
+      "lm_loss": 0.002310896525159478,
+      "order_loss": 51.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.189810752868652
+    },
+    {
+      "global_step": 2960,
+      "optimizer_step": 370,
+      "lr": 6.012640767930217e-06,
+      "lm_loss": 0.018192529678344727,
+      "order_loss": 54.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.424442291259766
+    },
+    {
+      "global_step": 2980,
+      "optimizer_step": 372,
+      "lr": 5.7864297241808e-06,
+      "lm_loss": 0.00038043089443817735,
+      "order_loss": 55.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.562880516052246
+    },
+    {
+      "global_step": 3000,
+      "optimizer_step": 375,
+      "lr": 5.454212938299255e-06,
+      "lm_loss": 0.0005834042676724494,
+      "order_loss": 41.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.094333171844482
+    },
+    {
+      "global_step": 3020,
+      "optimizer_step": 377,
+      "lr": 5.237531798197415e-06,
+      "lm_loss": 0.00628126272931695,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.97503137588501
+    },
+    {
+      "global_step": 3040,
+      "optimizer_step": 380,
+      "lr": 4.91981171298388e-06,
+      "lm_loss": 0.006694844923913479,
+      "order_loss": 51.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.162944793701172
+    },
+    {
+      "global_step": 3060,
+      "optimizer_step": 382,
+      "lr": 4.712926831217685e-06,
+      "lm_loss": 0.12685316801071167,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.095602989196777
+    },
+    {
+      "global_step": 3080,
+      "optimizer_step": 385,
+      "lr": 4.410094003202134e-06,
+      "lm_loss": 0.009770600125193596,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.509770393371582
+    },
+    {
+      "global_step": 3100,
+      "optimizer_step": 387,
+      "lr": 4.213259692436367e-06,
+      "lm_loss": 0.002140387427061796,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.970890522003174
+    },
+    {
+      "global_step": 3120,
+      "optimizer_step": 390,
+      "lr": 3.925686378029331e-06,
+      "lm_loss": 0.00836945790797472,
+      "order_loss": 56.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.695869445800781
+    },
+    {
+      "global_step": 3140,
+      "optimizer_step": 392,
+      "lr": 3.739144596292382e-06,
+      "lm_loss": 0.12146015465259552,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9027099609375
+    },
+    {
+      "global_step": 3160,
+      "optimizer_step": 395,
+      "lr": 3.4671842941897765e-06,
+      "lm_loss": 0.37217986583709717,
+      "order_loss": 71.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.528429985046387
+    },
+    {
+      "global_step": 3180,
+      "optimizer_step": 397,
+      "lr": 3.2911643474473646e-06,
+      "lm_loss": 0.00539372256025672,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.6303935050964355
+    },
+    {
+      "global_step": 3200,
+      "optimizer_step": 400,
+      "lr": 3.035151364093214e-06,
+      "lm_loss": 0.0002799053327180445,
+      "order_loss": 60.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.000279903411865
+    },
+    {
+      "global_step": 3220,
+      "optimizer_step": 402,
+      "lr": 2.869869624374699e-06,
+      "lm_loss": 0.13275915384292603,
+      "order_loss": 44.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.570259094238281
+    },
+    {
+      "global_step": 3240,
+      "optimizer_step": 405,
+      "lr": 2.6301186630158485e-06,
+      "lm_loss": 0.006312482059001923,
+      "order_loss": 62.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.193812370300293
+    },
+    {
+      "global_step": 3260,
+      "optimizer_step": 407,
+      "lr": 2.475778302439524e-06,
+      "lm_loss": 0.048964012414216995,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.830214023590088
+    },
+    {
+      "global_step": 3280,
+      "optimizer_step": 410,
+      "lr": 2.2525840762775863e-06,
+      "lm_loss": 0.2303650975227356,
+      "order_loss": 59.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.16786527633667
+    },
+    {
+      "global_step": 3300,
+      "optimizer_step": 412,
+      "lr": 2.1093748173018848e-06,
+      "lm_loss": 0.007891875691711903,
+      "order_loss": 58.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.851641654968262
+    },
+    {
+      "global_step": 3320,
+      "optimizer_step": 415,
+      "lr": 1.9030116872178316e-06,
+      "lm_loss": 0.013733092695474625,
+      "order_loss": 68.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.85748291015625
+    },
+    {
+      "global_step": 3340,
+      "optimizer_step": 417,
+      "lr": 1.771109569425547e-06,
+      "lm_loss": 0.021751023828983307,
+      "order_loss": 63.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.3342509269714355
+    },
+    {
+      "global_step": 3360,
+      "optimizer_step": 420,
+      "lr": 1.5818312067223672e-06,
+      "lm_loss": 0.12683044373989105,
+      "order_loss": 44.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.564330577850342
+    },
+    {
+      "global_step": 3380,
+      "optimizer_step": 422,
+      "lr": 1.4613983704244826e-06,
+      "lm_loss": 0.03174825385212898,
+      "order_loss": 67.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.719248294830322
+    },
+    {
+      "global_step": 3400,
+      "optimizer_step": 425,
+      "lr": 1.2894374450024338e-06,
+      "lm_loss": 0.009823920205235481,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.791073799133301
+    },
+    {
+      "global_step": 3420,
+      "optimizer_step": 427,
+      "lr": 1.180621931927592e-06,
+      "lm_loss": 0.010670330375432968,
+      "order_loss": 45.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.573170185089111
+    },
+    {
+      "global_step": 3440,
+      "optimizer_step": 430,
+      "lr": 1.0261898262753812e-06,
+      "lm_loss": 0.003456930862739682,
+      "order_loss": 64.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.409707069396973
+    },
+    {
+      "global_step": 3460,
+      "optimizer_step": 432,
+      "lr": 9.291253975900139e-07,
+      "lm_loss": 0.007835650816559792,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.976585865020752
+    },
+    {
+      "global_step": 3480,
+      "optimizer_step": 435,
+      "lr": 7.924119469434665e-07,
+      "lm_loss": 0.00920460931956768,
+      "order_loss": 43.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.321704387664795
+    },
+    {
+      "global_step": 3500,
+      "optimizer_step": 437,
+      "lr": 7.072179188262251e-07,
+      "lm_loss": 0.15896598994731903,
+      "order_loss": 59.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.096466064453125
+    },
+    {
+      "global_step": 3520,
+      "optimizer_step": 440,
+      "lr": 5.883911778138324e-07,
+      "lm_loss": 0.11559732258319855,
+      "order_loss": 51.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.209347248077393
+    },
+    {
+      "global_step": 3540,
+      "optimizer_step": 442,
+      "lr": 5.151722747865434e-07,
+      "lm_loss": 0.548110842704773,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.1731109619140625
+    },
+    {
+      "global_step": 3560,
+      "optimizer_step": 445,
+      "lr": 4.1437831084878974e-07,
+      "lm_loss": 0.2126510888338089,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.150151252746582
+    },
+    {
+      "global_step": 3580,
+      "optimizer_step": 447,
+      "lr": 3.5322453704410286e-07,
+      "lm_loss": 0.0028603943064808846,
+      "order_loss": 50.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.0653605461120605
+    },
+    {
+      "global_step": 3600,
+      "optimizer_step": 450,
+      "lr": 2.705872508804747e-07,
+      "lm_loss": 0.0031662171240895987,
+      "order_loss": 64.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.409416198730469
+    },
+    {
+      "global_step": 3620,
+      "optimizer_step": 452,
+      "lr": 2.215737794045436e-07,
+      "lm_loss": 0.052710987627506256,
+      "order_loss": 62.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.333961009979248
+    },
+    {
+      "global_step": 3640,
+      "optimizer_step": 455,
+      "lr": 1.571947526689349e-07,
+      "lm_loss": 0.008811240084469318,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.6650614738464355
+    },
+    {
+      "global_step": 3660,
+      "optimizer_step": 457,
+      "lr": 1.2038183319507955e-07,
+      "lm_loss": 0.005917538423091173,
+      "order_loss": 33.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 3.380917549133301
+    },
+    {
+      "global_step": 3680,
+      "optimizer_step": 460,
+      "lr": 7.434020362682704e-08,
+      "lm_loss": 0.032587382942438126,
+      "order_loss": 55.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.53258752822876
+    },
+    {
+      "global_step": 3700,
+      "optimizer_step": 462,
+      "lr": 4.977308833379324e-08,
+      "lm_loss": 0.0004790124949067831,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.937979221343994
+    },
+    {
+      "global_step": 3720,
+      "optimizer_step": 465,
+      "lr": 2.2125452477828047e-08,
+      "lm_loss": 0.01606699638068676,
+      "order_loss": 38.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 3.8598170280456543
+    },
+    {
+      "global_step": 3740,
+      "optimizer_step": 467,
+      "lr": 9.834340423678368e-09,
+      "lm_loss": 0.03486189246177673,
+      "order_loss": 50.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.034862041473389
+    }
+  ]
+}

cls_gemma_orderkd_last_l01/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_last_l01/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_last_l01/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_orderkd_last_l01/checkpoint-200/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_last_l01/checkpoint-200/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_last_l02/checkpoint-400/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_mid_l01/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_mid_l01/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_mid_l01/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_orderkd_mid_l01/checkpoint-200/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_mid_l01/checkpoint-200/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_mid_l01/checkpoint-200/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_orderkd_mid_l01/checkpoint-400/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/gemma-3-1b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/gemma-3-1b-it
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_gemma_orderkd_mid_l01/checkpoint-400/adapter_config.json

@@ -0,0 +1,44 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Gemma3ForCausalLM",
+    "parent_library": "transformers.models.gemma3.modeling_gemma3",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/gemma-3-1b-it",
+  "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": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_proj|gate_proj))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|o_proj|up_proj|down_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
+}

cls_gemma_orderkd_mid_l01/checkpoint-400/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

cls_gemma_orderkd_mid_l01/train_metrics.json

@@ -0,0 +1,1874 @@
+{
+  "loss_log": [
+    {
+      "global_step": 20,
+      "optimizer_step": 2,
+      "lr": 4.347826086956522e-06,
+      "lm_loss": 3.1184232234954834,
+      "order_loss": 63.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 9.462173461914062
+    },
+    {
+      "global_step": 40,
+      "optimizer_step": 5,
+      "lr": 1.0869565217391305e-05,
+      "lm_loss": 1.0916705131530762,
+      "order_loss": 56.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.747920513153076
+    },
+    {
+      "global_step": 60,
+      "optimizer_step": 7,
+      "lr": 1.5217391304347828e-05,
+      "lm_loss": 0.6041339039802551,
+      "order_loss": 42.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.8228840827941895
+    },
+    {
+      "global_step": 80,
+      "optimizer_step": 10,
+      "lr": 2.173913043478261e-05,
+      "lm_loss": 1.8147910833358765,
+      "order_loss": 55.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.346041202545166
+    },
+    {
+      "global_step": 100,
+      "optimizer_step": 12,
+      "lr": 2.608695652173913e-05,
+      "lm_loss": 0.5634512305259705,
+      "order_loss": 43.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.907201290130615
+    },
+    {
+      "global_step": 120,
+      "optimizer_step": 15,
+      "lr": 3.260869565217392e-05,
+      "lm_loss": 2.721200942993164,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 8.158700942993164
+    },
+    {
+      "global_step": 140,
+      "optimizer_step": 17,
+      "lr": 3.695652173913043e-05,
+      "lm_loss": 0.6781831979751587,
+      "order_loss": 54.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.084433078765869
+    },
+    {
+      "global_step": 160,
+      "optimizer_step": 20,
+      "lr": 4.347826086956522e-05,
+      "lm_loss": 0.41744333505630493,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.07369327545166
+    },
+    {
+      "global_step": 180,
+      "optimizer_step": 22,
+      "lr": 4.782608695652174e-05,
+      "lm_loss": 0.06321373581886292,
+      "order_loss": 54.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.50071382522583
+    },
+    {
+      "global_step": 200,
+      "optimizer_step": 25,
+      "lr": 4.9997541293989384e-05,
+      "lm_loss": 0.02435319498181343,
+      "order_loss": 54.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.46185302734375
+    },
+    {
+      "global_step": 220,
+      "optimizer_step": 27,
+      "lr": 4.999016565957633e-05,
+      "lm_loss": 0.019743939861655235,
+      "order_loss": 61.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.144743919372559
+    },
+    {
+      "global_step": 240,
+      "optimizer_step": 30,
+      "lr": 4.996988640512931e-05,
+      "lm_loss": 0.3459993302822113,
+      "order_loss": 39.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.314749240875244
+    },
+    {
+      "global_step": 260,
+      "optimizer_step": 32,
+      "lr": 4.995022691166621e-05,
+      "lm_loss": 0.0298320259898901,
+      "order_loss": 62.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.248581886291504
+    },
+    {
+      "global_step": 280,
+      "optimizer_step": 35,
+      "lr": 4.991153735294049e-05,
+      "lm_loss": 0.46130573749542236,
+      "order_loss": 60.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.523805618286133
+    },
+    {
+      "global_step": 300,
+      "optimizer_step": 37,
+      "lr": 4.987961816680492e-05,
+      "lm_loss": 0.1003972664475441,
+      "order_loss": 54.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.569147109985352
+    },
+    {
+      "global_step": 320,
+      "optimizer_step": 40,
+      "lr": 4.9822565862832806e-05,
+      "lm_loss": 0.13187472522258759,
+      "order_loss": 40.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.1943745613098145
+    },
+    {
+      "global_step": 340,
+      "optimizer_step": 42,
+      "lr": 4.9778426220595456e-05,
+      "lm_loss": 0.32335221767425537,
+      "order_loss": 50.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.323352336883545
+    },
+    {
+      "global_step": 360,
+      "optimizer_step": 45,
+      "lr": 4.970308130276272e-05,
+      "lm_loss": 0.7211238741874695,
+      "order_loss": 56.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.377373695373535
+    },
+    {
+      "global_step": 380,
+      "optimizer_step": 47,
+      "lr": 4.96467754629559e-05,
+      "lm_loss": 0.03598100692033768,
+      "order_loss": 41.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.223481178283691
+    },
+    {
+      "global_step": 400,
+      "optimizer_step": 50,
+      "lr": 4.955323054879298e-05,
+      "lm_loss": 0.10630018264055252,
+      "order_loss": 47.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.825050354003906
+    },
+    {
+      "global_step": 420,
+      "optimizer_step": 52,
+      "lr": 4.948482772521345e-05,
+      "lm_loss": 0.11145680397748947,
+      "order_loss": 66.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.705206871032715
+    },
+    {
+      "global_step": 440,
+      "optimizer_step": 55,
+      "lr": 4.937319780454559e-05,
+      "lm_loss": 0.18087011575698853,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.618370056152344
+    },
+    {
+      "global_step": 460,
+      "optimizer_step": 57,
+      "lr": 4.929278208117378e-05,
+      "lm_loss": 0.6563955545425415,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.625145435333252
+    },
+    {
+      "global_step": 480,
+      "optimizer_step": 60,
+      "lr": 4.916320437477008e-05,
+      "lm_loss": 0.32431942224502563,
+      "order_loss": 57.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.043069362640381
+    },
+    {
+      "global_step": 500,
+      "optimizer_step": 62,
+      "lr": 4.907087460240999e-05,
+      "lm_loss": 0.14991332590579987,
+      "order_loss": 61.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.274913311004639
+    },
+    {
+      "global_step": 520,
+      "optimizer_step": 65,
+      "lr": 4.892350839330522e-05,
+      "lm_loss": 0.1010974645614624,
+      "order_loss": 50.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.163597583770752
+    },
+    {
+      "global_step": 540,
+      "optimizer_step": 67,
+      "lr": 4.881937806807241e-05,
+      "lm_loss": 0.15496166050434113,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.779961585998535
+    },
+    {
+      "global_step": 560,
+      "optimizer_step": 70,
+      "lr": 4.865440450576874e-05,
+      "lm_loss": 0.026232725009322166,
+      "order_loss": 56.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.713732719421387
+    },
+    {
+      "global_step": 580,
+      "optimizer_step": 72,
+      "lr": 4.853860162957552e-05,
+      "lm_loss": 0.09959332644939423,
+      "order_loss": 57.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.787093162536621
+    },
+    {
+      "global_step": 600,
+      "optimizer_step": 75,
+      "lr": 4.8356223507364996e-05,
+      "lm_loss": 0.01794126257300377,
+      "order_loss": 52.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.299191474914551
+    },
+    {
+      "global_step": 620,
+      "optimizer_step": 77,
+      "lr": 4.822889043057446e-05,
+      "lm_loss": 0.26736268401145935,
+      "order_loss": 49.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.204862594604492
+    },
+    {
+      "global_step": 640,
+      "optimizer_step": 80,
+      "lr": 4.802933193625593e-05,
+      "lm_loss": 0.007746524177491665,
+      "order_loss": 67.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.695246696472168
+    },
+    {
+      "global_step": 660,
+      "optimizer_step": 82,
+      "lr": 4.789062518269812e-05,
+      "lm_loss": 0.03736429661512375,
+      "order_loss": 52.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.318614482879639
+    },
+    {
+      "global_step": 680,
+      "optimizer_step": 85,
+      "lr": 4.767413162299501e-05,
+      "lm_loss": 0.06276468187570572,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.031514644622803
+    },
+    {
+      "global_step": 700,
+      "optimizer_step": 87,
+      "lr": 4.752422169756048e-05,
+      "lm_loss": 0.22046461701393127,
+      "order_loss": 67.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.970464706420898
+    },
+    {
+      "global_step": 720,
+      "optimizer_step": 90,
+      "lr": 4.729105919657818e-05,
+      "lm_loss": 0.4013015925884247,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.151301383972168
+    },
+    {
+      "global_step": 740,
+      "optimizer_step": 92,
+      "lr": 4.7130130375625305e-05,
+      "lm_loss": 0.15300190448760986,
+      "order_loss": 55.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.65300178527832
+    },
+    {
+      "global_step": 760,
+      "optimizer_step": 95,
+      "lr": 4.6880585547718845e-05,
+      "lm_loss": 0.3297501802444458,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.079750061035156
+    },
+    {
+      "global_step": 780,
+      "optimizer_step": 97,
+      "lr": 4.670883565255264e-05,
+      "lm_loss": 0.017043394967913628,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.267043590545654
+    },
+    {
+      "global_step": 800,
+      "optimizer_step": 100,
+      "lr": 4.6443215250006806e-05,
+      "lm_loss": 0.3523687720298767,
+      "order_loss": 56.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9461188316345215
+    },
+    {
+      "global_step": 820,
+      "optimizer_step": 102,
+      "lr": 4.626085540370762e-05,
+      "lm_loss": 0.012201391160488129,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.793451309204102
+    },
+    {
+      "global_step": 840,
+      "optimizer_step": 105,
+      "lr": 4.597948593966256e-05,
+      "lm_loss": 0.07073015719652176,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.039480209350586
+    },
+    {
+      "global_step": 860,
+      "optimizer_step": 107,
+      "lr": 4.5786740307563636e-05,
+      "lm_loss": 0.0668209120631218,
+      "order_loss": 63.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.441821098327637
+    },
+    {
+      "global_step": 880,
+      "optimizer_step": 110,
+      "lr": 4.5489967654649474e-05,
+      "lm_loss": 0.1039772778749466,
+      "order_loss": 63.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.416477203369141
+    },
+    {
+      "global_step": 900,
+      "optimizer_step": 112,
+      "lr": 4.528707316878232e-05,
+      "lm_loss": 0.1377195566892624,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.075219631195068
+    },
+    {
+      "global_step": 920,
+      "optimizer_step": 115,
+      "lr": 4.497526213395623e-05,
+      "lm_loss": 0.1522826999425888,
+      "order_loss": 55.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.71478271484375
+    },
+    {
+      "global_step": 940,
+      "optimizer_step": 117,
+      "lr": 4.476246820180259e-05,
+      "lm_loss": 0.6368707418441772,
+      "order_loss": 52.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.855620861053467
+    },
+    {
+      "global_step": 960,
+      "optimizer_step": 120,
+      "lr": 4.443600207791081e-05,
+      "lm_loss": 0.2829626798629761,
+      "order_loss": 61.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.439212799072266
+    },
+    {
+      "global_step": 980,
+      "optimizer_step": 122,
+      "lr": 4.421357027581921e-05,
+      "lm_loss": 0.8739344477653503,
+      "order_loss": 55.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.405184268951416
+    },
+    {
+      "global_step": 1000,
+      "optimizer_step": 125,
+      "lr": 4.38728503704354e-05,
+      "lm_loss": 0.07859532535076141,
+      "order_loss": 55.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.641095161437988
+    },
+    {
+      "global_step": 1020,
+      "optimizer_step": 127,
+      "lr": 4.364105412207914e-05,
+      "lm_loss": 0.25118908286094666,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.751189231872559
+    },
+    {
+      "global_step": 1040,
+      "optimizer_step": 130,
+      "lr": 4.328649926419819e-05,
+      "lm_loss": 0.06467773020267487,
+      "order_loss": 52.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.283427715301514
+    },
+    {
+      "global_step": 1060,
+      "optimizer_step": 132,
+      "lr": 4.304562350447008e-05,
+      "lm_loss": 0.18095380067825317,
+      "order_loss": 56.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.7747039794921875
+    },
+    {
+      "global_step": 1080,
+      "optimizer_step": 135,
+      "lr": 4.267766952966369e-05,
+      "lm_loss": 0.02313944138586521,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.273139476776123
+    },
+    {
+      "global_step": 1100,
+      "optimizer_step": 137,
+      "lr": 4.242801035442058e-05,
+      "lm_loss": 0.5191475749015808,
+      "order_loss": 62.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.7066473960876465
+    },
+    {
+      "global_step": 1120,
+      "optimizer_step": 140,
+      "lr": 4.204710956908763e-05,
+      "lm_loss": 0.81064772605896,
+      "order_loss": 62.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.998147964477539
+    },
+    {
+      "global_step": 1140,
+      "optimizer_step": 142,
+      "lr": 4.178897387117546e-05,
+      "lm_loss": 0.0647754892706871,
+      "order_loss": 61.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.158525466918945
+    },
+    {
+      "global_step": 1160,
+      "optimizer_step": 145,
+      "lr": 4.139559449654561e-05,
+      "lm_loss": 0.048019688576459885,
+      "order_loss": 53.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.391769886016846
+    },
+    {
+      "global_step": 1180,
+      "optimizer_step": 147,
+      "lr": 4.1129299588552193e-05,
+      "lm_loss": 0.160772904753685,
+      "order_loss": 61.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.285772800445557
+    },
+    {
+      "global_step": 1200,
+      "optimizer_step": 150,
+      "lr": 4.072392518512623e-05,
+      "lm_loss": 0.4824674129486084,
+      "order_loss": 63.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.8262176513671875
+    },
+    {
+      "global_step": 1220,
+      "optimizer_step": 152,
+      "lr": 4.0449798409325665e-05,
+      "lm_loss": 0.05526652932167053,
+      "order_loss": 52.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.305266380310059
+    },
+    {
+      "global_step": 1240,
+      "optimizer_step": 155,
+      "lr": 4.0032927282460146e-05,
+      "lm_loss": 0.33561477065086365,
+      "order_loss": 60.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.3668646812438965
+    },
+    {
+      "global_step": 1260,
+      "optimizer_step": 157,
+      "lr": 3.9751305608428205e-05,
+      "lm_loss": 0.17131206393241882,
+      "order_loss": 57.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.858811855316162
+    },
+    {
+      "global_step": 1280,
+      "optimizer_step": 160,
+      "lr": 3.932345019579495e-05,
+      "lm_loss": 0.04662935063242912,
+      "order_loss": 61.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.234129428863525
+    },
+    {
+      "global_step": 1300,
+      "optimizer_step": 162,
+      "lr": 3.90346798061902e-05,
+      "lm_loss": 0.050219446420669556,
+      "order_loss": 50.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.081469535827637
+    },
+    {
+      "global_step": 1320,
+      "optimizer_step": 165,
+      "lr": 3.859636604786372e-05,
+      "lm_loss": 0.34347474575042725,
+      "order_loss": 48.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.155974864959717
+    },
+    {
+      "global_step": 1340,
+      "optimizer_step": 167,
+      "lr": 3.830080191288342e-05,
+      "lm_loss": 0.4303498864173889,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.867849826812744
+    },
+    {
+      "global_step": 1360,
+      "optimizer_step": 170,
+      "lr": 3.785256860483054e-05,
+      "lm_loss": 0.07431875914335251,
+      "order_loss": 53.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.418068885803223
+    },
+    {
+      "global_step": 1380,
+      "optimizer_step": 172,
+      "lr": 3.755057404586455e-05,
+      "lm_loss": 0.14596708118915558,
+      "order_loss": 39.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.06784200668335
+    },
+    {
+      "global_step": 1400,
+      "optimizer_step": 175,
+      "lr": 3.7092972177631e-05,
+      "lm_loss": 0.13190682232379913,
+      "order_loss": 65.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.631906986236572
+    },
+    {
+      "global_step": 1420,
+      "optimizer_step": 177,
+      "lr": 3.678491842064995e-05,
+      "lm_loss": 0.15802578628063202,
+      "order_loss": 51.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.283025741577148
+    },
+    {
+      "global_step": 1440,
+      "optimizer_step": 180,
+      "lr": 3.631851049805793e-05,
+      "lm_loss": 0.30685389041900635,
+      "order_loss": 45.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.869353771209717
+    },
+    {
+      "global_step": 1460,
+      "optimizer_step": 182,
+      "lr": 3.600477621728473e-05,
+      "lm_loss": 0.01724999025464058,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.642250061035156
+    },
+    {
+      "global_step": 1480,
+      "optimizer_step": 185,
+      "lr": 3.5530135570974274e-05,
+      "lm_loss": 0.25045639276504517,
+      "order_loss": 50.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.3129563331604
+    },
+    {
+      "global_step": 1500,
+      "optimizer_step": 187,
+      "lr": 3.521110642339991e-05,
+      "lm_loss": 0.4265330731868744,
+      "order_loss": 52.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.707783222198486
+    },
+    {
+      "global_step": 1520,
+      "optimizer_step": 190,
+      "lr": 3.4728816504063726e-05,
+      "lm_loss": 0.23969930410385132,
+      "order_loss": 61.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.427199363708496
+    },
+    {
+      "global_step": 1540,
+      "optimizer_step": 192,
+      "lr": 3.440488465537943e-05,
+      "lm_loss": 0.006110867951065302,
+      "order_loss": 65.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.506110668182373
+    },
+    {
+      "global_step": 1560,
+      "optimizer_step": 195,
+      "lr": 3.391553831655782e-05,
+      "lm_loss": 0.01415237132459879,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.951652526855469
+    },
+    {
+      "global_step": 1580,
+      "optimizer_step": 197,
+      "lr": 3.358710195908653e-05,
+      "lm_loss": 0.2871032953262329,
+      "order_loss": 70.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.287103176116943
+    },
+    {
+      "global_step": 1600,
+      "optimizer_step": 200,
+      "lr": 3.309130072840379e-05,
+      "lm_loss": 0.23844397068023682,
+      "order_loss": 58.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.050943851470947
+    },
+    {
+      "global_step": 1620,
+      "optimizer_step": 202,
+      "lr": 3.2758763591623295e-05,
+      "lm_loss": 0.1546662300825119,
+      "order_loss": 50.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.185916423797607
+    },
+    {
+      "global_step": 1640,
+      "optimizer_step": 205,
+      "lr": 3.225711693136156e-05,
+      "lm_loss": 0.08392491936683655,
+      "order_loss": 60.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.146424770355225
+    },
+    {
+      "global_step": 1660,
+      "optimizer_step": 207,
+      "lr": 3.1920887785621235e-05,
+      "lm_loss": 0.043332748115062714,
+      "order_loss": 61.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.230832576751709
+    },
+    {
+      "global_step": 1680,
+      "optimizer_step": 210,
+      "lr": 3.141401234354067e-05,
+      "lm_loss": 0.13834291696548462,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.10709285736084
+    },
+    {
+      "global_step": 1700,
+      "optimizer_step": 212,
+      "lr": 3.10745044975816e-05,
+      "lm_loss": 0.2134607881307602,
+      "order_loss": 50.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.275960922241211
+    },
+    {
+      "global_step": 1720,
+      "optimizer_step": 215,
+      "lr": 3.056302334890786e-05,
+      "lm_loss": 0.01857478730380535,
+      "order_loss": 74.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.4248247146606445
+    },
+    {
+      "global_step": 1740,
+      "optimizer_step": 217,
+      "lr": 3.022065414180425e-05,
+      "lm_loss": 0.006793432869017124,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.25679349899292
+    },
+    {
+      "global_step": 1760,
+      "optimizer_step": 220,
+      "lr": 2.9705196023315006e-05,
+      "lm_loss": 0.013562669977545738,
+      "order_loss": 59.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9510626792907715
+    },
+    {
+      "global_step": 1780,
+      "optimizer_step": 222,
+      "lr": 2.9360386311461275e-05,
+      "lm_loss": 0.0029967674054205418,
+      "order_loss": 47.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.721746921539307
+    },
+    {
+      "global_step": 1800,
+      "optimizer_step": 225,
+      "lr": 2.884158484861325e-05,
+      "lm_loss": 0.012671629898250103,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.76267147064209
+    },
+    {
+      "global_step": 1820,
+      "optimizer_step": 227,
+      "lr": 2.849475848838749e-05,
+      "lm_loss": 0.0835554227232933,
+      "order_loss": 39.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.036680221557617
+    },
+    {
+      "global_step": 1840,
+      "optimizer_step": 230,
+      "lr": 2.7973251416434175e-05,
+      "lm_loss": 0.05105651170015335,
+      "order_loss": 29.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 3.0041816234588623
+    },
+    {
+      "global_step": 1860,
+      "optimizer_step": 232,
+      "lr": 2.7624834743173867e-05,
+      "lm_loss": 0.5029534101486206,
+      "order_loss": 75.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 8.00295352935791
+    },
+    {
+      "global_step": 1880,
+      "optimizer_step": 235,
+      "lr": 2.710126312323119e-05,
+      "lm_loss": 0.12120761722326279,
+      "order_loss": 62.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.3712077140808105
+    },
+    {
+      "global_step": 1900,
+      "optimizer_step": 237,
+      "lr": 2.6751684427161683e-05,
+      "lm_loss": 0.17923295497894287,
+      "order_loss": 40.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.210483074188232
+    },
+    {
+      "global_step": 1920,
+      "optimizer_step": 240,
+      "lr": 2.6226691858185454e-05,
+      "lm_loss": 0.004263256676495075,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.504263401031494
+    },
+    {
+      "global_step": 1940,
+      "optimizer_step": 242,
+      "lr": 2.587638085794541e-05,
+      "lm_loss": 0.04619055613875389,
+      "order_loss": 56.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.7024407386779785
+    },
+    {
+      "global_step": 1960,
+      "optimizer_step": 245,
+      "lr": 2.535061268558906e-05,
+      "lm_loss": 0.11890175938606262,
+      "order_loss": 71.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.27515172958374
+    },
+    {
+      "global_step": 1980,
+      "optimizer_step": 247,
+      "lr": 2.5e-05,
+      "lm_loss": 0.25617969036102295,
+      "order_loss": 47.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.9436798095703125
+    },
+    {
+      "global_step": 2000,
+      "optimizer_step": 250,
+      "lr": 2.447410252332528e-05,
+      "lm_loss": 0.013576528988778591,
+      "order_loss": 54.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.451076507568359
+    },
+    {
+      "global_step": 2020,
+      "optimizer_step": 252,
+      "lr": 2.4123619142054597e-05,
+      "lm_loss": 0.04630662128329277,
+      "order_loss": 50.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.046306610107422
+    },
+    {
+      "global_step": 2040,
+      "optimizer_step": 255,
+      "lr": 2.3598238819070202e-05,
+      "lm_loss": 0.14818225800991058,
+      "order_loss": 61.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.304432392120361
+    },
+    {
+      "global_step": 2060,
+      "optimizer_step": 257,
+      "lr": 2.3248315572838316e-05,
+      "lm_loss": 0.0021958579309284687,
+      "order_loss": 44.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.470945835113525
+    },
+    {
+      "global_step": 2080,
+      "optimizer_step": 260,
+      "lr": 2.272409822584325e-05,
+      "lm_loss": 0.002902108011767268,
+      "order_loss": 43.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.377902030944824
+    },
+    {
+      "global_step": 2100,
+      "optimizer_step": 262,
+      "lr": 2.237516525682614e-05,
+      "lm_loss": 0.016654551029205322,
+      "order_loss": 47.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.7041544914245605
+    },
+    {
+      "global_step": 2120,
+      "optimizer_step": 265,
+      "lr": 2.185275527853437e-05,
+      "lm_loss": 0.012752072885632515,
+      "order_loss": 43.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.325252056121826
+    },
+    {
+      "global_step": 2140,
+      "optimizer_step": 267,
+      "lr": 2.1505241511612522e-05,
+      "lm_loss": 0.009639913216233253,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.947139739990234
+    },
+    {
+      "global_step": 2160,
+      "optimizer_step": 270,
+      "lr": 2.0985281073035076e-05,
+      "lm_loss": 0.11398650705814362,
+      "order_loss": 44.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.551486492156982
+    },
+    {
+      "global_step": 2180,
+      "optimizer_step": 272,
+      "lr": 2.0639613688538734e-05,
+      "lm_loss": 0.0007875743904151022,
+      "order_loss": 51.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.094537734985352
+    },
+    {
+      "global_step": 2200,
+      "optimizer_step": 275,
+      "lr": 2.0122741949596797e-05,
+      "lm_loss": 0.009382687509059906,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.759382724761963
+    },
+    {
+      "global_step": 2220,
+      "optimizer_step": 277,
+      "lr": 1.977934585819576e-05,
+      "lm_loss": 0.2916325628757477,
+      "order_loss": 48.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.104132652282715
+    },
+    {
+      "global_step": 2240,
+      "optimizer_step": 280,
+      "lr": 1.9266198182035058e-05,
+      "lm_loss": 0.006090455688536167,
+      "order_loss": 61.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.1623406410217285
+    },
+    {
+      "global_step": 2260,
+      "optimizer_step": 282,
+      "lr": 1.8925495502418406e-05,
+      "lm_loss": 0.06058869510889053,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.685588836669922
+    },
+    {
+      "global_step": 2280,
+      "optimizer_step": 285,
+      "lr": 1.841670267439088e-05,
+      "lm_loss": 0.0836532935500145,
+      "order_loss": 51.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.271153450012207
+    },
+    {
+      "global_step": 2300,
+      "optimizer_step": 287,
+      "lr": 1.8079112214378768e-05,
+      "lm_loss": 0.05897977575659752,
+      "order_loss": 58.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9339799880981445
+    },
+    {
+      "global_step": 2320,
+      "optimizer_step": 290,
+      "lr": 1.757529966665141e-05,
+      "lm_loss": 0.1328563094139099,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.789106369018555
+    },
+    {
+      "global_step": 2340,
+      "optimizer_step": 292,
+      "lr": 1.72412364083767e-05,
+      "lm_loss": 0.044375427067279816,
+      "order_loss": 63.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.356875419616699
+    },
+    {
+      "global_step": 2360,
+      "optimizer_step": 295,
+      "lr": 1.6743023451120832e-05,
+      "lm_loss": 0.18005213141441345,
+      "order_loss": 58.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.992552280426025
+    },
+    {
+      "global_step": 2380,
+      "optimizer_step": 297,
+      "lr": 1.641289804091347e-05,
+      "lm_loss": 0.07155724614858627,
+      "order_loss": 48.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.88405704498291
+    },
+    {
+      "global_step": 2400,
+      "optimizer_step": 300,
+      "lr": 1.5920897101019383e-05,
+      "lm_loss": 0.21017329394817352,
+      "order_loss": 48.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.0226731300354
+    },
+    {
+      "global_step": 2420,
+      "optimizer_step": 302,
+      "lr": 1.5595115344620576e-05,
+      "lm_loss": 0.013050842098891735,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.7630510330200195
+    },
+    {
+      "global_step": 2440,
+      "optimizer_step": 305,
+      "lr": 1.5109931212873534e-05,
+      "lm_loss": 0.022372275590896606,
+      "order_loss": 61.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.116122245788574
+    },
+    {
+      "global_step": 2460,
+      "optimizer_step": 307,
+      "lr": 1.4788893576600099e-05,
+      "lm_loss": 0.2720758318901062,
+      "order_loss": 64.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.709575653076172
+    },
+    {
+      "global_step": 2480,
+      "optimizer_step": 310,
+      "lr": 1.4311122664242954e-05,
+      "lm_loss": 0.0060526239685714245,
+      "order_loss": 51.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.162302494049072
+    },
+    {
+      "global_step": 2500,
+      "optimizer_step": 312,
+      "lr": 1.399522378271527e-05,
+      "lm_loss": 0.8424654603004456,
+      "order_loss": 57.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.529965400695801
+    },
+    {
+      "global_step": 2520,
+      "optimizer_step": 315,
+      "lr": 1.3525453388311554e-05,
+      "lm_loss": 0.12010786682367325,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.057607650756836
+    },
+    {
+      "global_step": 2540,
+      "optimizer_step": 317,
+      "lr": 1.3215081579350058e-05,
+      "lm_loss": 0.6840029954910278,
+      "order_loss": 47.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.371502876281738
+    },
+    {
+      "global_step": 2560,
+      "optimizer_step": 320,
+      "lr": 1.2753889166848909e-05,
+      "lm_loss": 0.005617386195808649,
+      "order_loss": 63.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.318117618560791
+    },
+    {
+      "global_step": 2580,
+      "optimizer_step": 322,
+      "lr": 1.2449425954135452e-05,
+      "lm_loss": 0.009596875868737698,
+      "order_loss": 46.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.603346824645996
+    },
+    {
+      "global_step": 2600,
+      "optimizer_step": 325,
+      "lr": 1.1997378443025634e-05,
+      "lm_loss": 0.006365991663187742,
+      "order_loss": 50.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.068865776062012
+    },
+    {
+      "global_step": 2620,
+      "optimizer_step": 327,
+      "lr": 1.1699198087116589e-05,
+      "lm_loss": 0.0015480580041185021,
+      "order_loss": 47.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.7515482902526855
+    },
+    {
+      "global_step": 2640,
+      "optimizer_step": 330,
+      "lr": 1.1256851155542298e-05,
+      "lm_loss": 0.0021843877620995045,
+      "order_loss": 56.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.6896843910217285
+    },
+    {
+      "global_step": 2660,
+      "optimizer_step": 332,
+      "lr": 1.0965320193809808e-05,
+      "lm_loss": 0.01388256810605526,
+      "order_loss": 52.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.263882637023926
+    },
+    {
+      "global_step": 2680,
+      "optimizer_step": 335,
+      "lr": 1.0533217595504858e-05,
+      "lm_loss": 0.0009903698228299618,
+      "order_loss": 53.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.344740390777588
+    },
+    {
+      "global_step": 2700,
+      "optimizer_step": 337,
+      "lr": 1.0248694391571801e-05,
+      "lm_loss": 0.0581960529088974,
+      "order_loss": 51.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.183196067810059
+    },
+    {
+      "global_step": 2720,
+      "optimizer_step": 340,
+      "lr": 9.82736728745186e-06,
+      "lm_loss": 0.05928004905581474,
+      "order_loss": 57.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.809279918670654
+    },
+    {
+      "global_step": 2740,
+      "optimizer_step": 342,
+      "lr": 9.550201590674343e-06,
+      "lm_loss": 0.00916436966508627,
+      "order_loss": 51.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.165414333343506
+    },
+    {
+      "global_step": 2760,
+      "optimizer_step": 345,
+      "lr": 9.140167895908867e-06,
+      "lm_loss": 0.02078745700418949,
+      "order_loss": 45.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.552037239074707
+    },
+    {
+      "global_step": 2780,
+      "optimizer_step": 347,
+      "lr": 8.870700411447816e-06,
+      "lm_loss": 0.0645354613661766,
+      "order_loss": 51.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.252035617828369
+    },
+    {
+      "global_step": 2800,
+      "optimizer_step": 350,
+      "lr": 8.472464158814255e-06,
+      "lm_loss": 0.03742121905088425,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.537421226501465
+    },
+    {
+      "global_step": 2820,
+      "optimizer_step": 352,
+      "lr": 8.211026128824539e-06,
+      "lm_loss": 0.004177003167569637,
+      "order_loss": 56.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.597927093505859
+    },
+    {
+      "global_step": 2840,
+      "optimizer_step": 355,
+      "lr": 7.825076849127458e-06,
+      "lm_loss": 0.017688076943159103,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.673938274383545
+    },
+    {
+      "global_step": 2860,
+      "optimizer_step": 357,
+      "lr": 7.571989645579419e-06,
+      "lm_loss": 0.0998288244009018,
+      "order_loss": 60.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.099828720092773
+    },
+    {
+      "global_step": 2880,
+      "optimizer_step": 360,
+      "lr": 7.198801765896115e-06,
+      "lm_loss": 0.24300594627857208,
+      "order_loss": 64.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.649255752563477
+    },
+    {
+      "global_step": 2900,
+      "optimizer_step": 362,
+      "lr": 6.9543764955299265e-06,
+      "lm_loss": 0.03150952234864235,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.9690093994140625
+    },
+    {
+      "global_step": 2920,
+      "optimizer_step": 365,
+      "lr": 6.594408756022272e-06,
+      "lm_loss": 0.1402754932641983,
+      "order_loss": 58.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.952775478363037
+    },
+    {
+      "global_step": 2940,
+      "optimizer_step": 367,
+      "lr": 6.358945877920861e-06,
+      "lm_loss": 0.002310896525159478,
+      "order_loss": 51.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.189810752868652
+    },
+    {
+      "global_step": 2960,
+      "optimizer_step": 370,
+      "lr": 6.012640767930217e-06,
+      "lm_loss": 0.018192529678344727,
+      "order_loss": 54.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.424442291259766
+    },
+    {
+      "global_step": 2980,
+      "optimizer_step": 372,
+      "lr": 5.7864297241808e-06,
+      "lm_loss": 0.00038043089443817735,
+      "order_loss": 55.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.562880516052246
+    },
+    {
+      "global_step": 3000,
+      "optimizer_step": 375,
+      "lr": 5.454212938299255e-06,
+      "lm_loss": 0.0005834042676724494,
+      "order_loss": 41.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.094333171844482
+    },
+    {
+      "global_step": 3020,
+      "optimizer_step": 377,
+      "lr": 5.237531798197415e-06,
+      "lm_loss": 0.00628126272931695,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.97503137588501
+    },
+    {
+      "global_step": 3040,
+      "optimizer_step": 380,
+      "lr": 4.91981171298388e-06,
+      "lm_loss": 0.006694844923913479,
+      "order_loss": 51.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.162944793701172
+    },
+    {
+      "global_step": 3060,
+      "optimizer_step": 382,
+      "lr": 4.712926831217685e-06,
+      "lm_loss": 0.12685316801071167,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.095602989196777
+    },
+    {
+      "global_step": 3080,
+      "optimizer_step": 385,
+      "lr": 4.410094003202134e-06,
+      "lm_loss": 0.009770600125193596,
+      "order_loss": 45.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.509770393371582
+    },
+    {
+      "global_step": 3100,
+      "optimizer_step": 387,
+      "lr": 4.213259692436367e-06,
+      "lm_loss": 0.002140387427061796,
+      "order_loss": 49.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.970890522003174
+    },
+    {
+      "global_step": 3120,
+      "optimizer_step": 390,
+      "lr": 3.925686378029331e-06,
+      "lm_loss": 0.00836945790797472,
+      "order_loss": 56.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.695869445800781
+    },
+    {
+      "global_step": 3140,
+      "optimizer_step": 392,
+      "lr": 3.739144596292382e-06,
+      "lm_loss": 0.12146015465259552,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.9027099609375
+    },
+    {
+      "global_step": 3160,
+      "optimizer_step": 395,
+      "lr": 3.4671842941897765e-06,
+      "lm_loss": 0.37217986583709717,
+      "order_loss": 71.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 7.528429985046387
+    },
+    {
+      "global_step": 3180,
+      "optimizer_step": 397,
+      "lr": 3.2911643474473646e-06,
+      "lm_loss": 0.00539372256025672,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.6303935050964355
+    },
+    {
+      "global_step": 3200,
+      "optimizer_step": 400,
+      "lr": 3.035151364093214e-06,
+      "lm_loss": 0.0002799053327180445,
+      "order_loss": 60.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.000279903411865
+    },
+    {
+      "global_step": 3220,
+      "optimizer_step": 402,
+      "lr": 2.869869624374699e-06,
+      "lm_loss": 0.13275915384292603,
+      "order_loss": 44.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.570259094238281
+    },
+    {
+      "global_step": 3240,
+      "optimizer_step": 405,
+      "lr": 2.6301186630158485e-06,
+      "lm_loss": 0.006312482059001923,
+      "order_loss": 62.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.193812370300293
+    },
+    {
+      "global_step": 3260,
+      "optimizer_step": 407,
+      "lr": 2.475778302439524e-06,
+      "lm_loss": 0.048964012414216995,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.830214023590088
+    },
+    {
+      "global_step": 3280,
+      "optimizer_step": 410,
+      "lr": 2.2525840762775863e-06,
+      "lm_loss": 0.2303650975227356,
+      "order_loss": 59.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.16786527633667
+    },
+    {
+      "global_step": 3300,
+      "optimizer_step": 412,
+      "lr": 2.1093748173018848e-06,
+      "lm_loss": 0.007891875691711903,
+      "order_loss": 58.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.851641654968262
+    },
+    {
+      "global_step": 3320,
+      "optimizer_step": 415,
+      "lr": 1.9030116872178316e-06,
+      "lm_loss": 0.013733092695474625,
+      "order_loss": 68.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.85748291015625
+    },
+    {
+      "global_step": 3340,
+      "optimizer_step": 417,
+      "lr": 1.771109569425547e-06,
+      "lm_loss": 0.021751023828983307,
+      "order_loss": 63.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.3342509269714355
+    },
+    {
+      "global_step": 3360,
+      "optimizer_step": 420,
+      "lr": 1.5818312067223672e-06,
+      "lm_loss": 0.12683044373989105,
+      "order_loss": 44.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.564330577850342
+    },
+    {
+      "global_step": 3380,
+      "optimizer_step": 422,
+      "lr": 1.4613983704244826e-06,
+      "lm_loss": 0.03174825385212898,
+      "order_loss": 67.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.719248294830322
+    },
+    {
+      "global_step": 3400,
+      "optimizer_step": 425,
+      "lr": 1.2894374450024338e-06,
+      "lm_loss": 0.009823920205235481,
+      "order_loss": 57.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.791073799133301
+    },
+    {
+      "global_step": 3420,
+      "optimizer_step": 427,
+      "lr": 1.180621931927592e-06,
+      "lm_loss": 0.010670330375432968,
+      "order_loss": 45.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.573170185089111
+    },
+    {
+      "global_step": 3440,
+      "optimizer_step": 430,
+      "lr": 1.0261898262753812e-06,
+      "lm_loss": 0.003456930862739682,
+      "order_loss": 64.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.409707069396973
+    },
+    {
+      "global_step": 3460,
+      "optimizer_step": 432,
+      "lr": 9.291253975900139e-07,
+      "lm_loss": 0.007835650816559792,
+      "order_loss": 59.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.976585865020752
+    },
+    {
+      "global_step": 3480,
+      "optimizer_step": 435,
+      "lr": 7.924119469434665e-07,
+      "lm_loss": 0.00920460931956768,
+      "order_loss": 43.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.321704387664795
+    },
+    {
+      "global_step": 3500,
+      "optimizer_step": 437,
+      "lr": 7.072179188262251e-07,
+      "lm_loss": 0.15896598994731903,
+      "order_loss": 59.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.096466064453125
+    },
+    {
+      "global_step": 3520,
+      "optimizer_step": 440,
+      "lr": 5.883911778138324e-07,
+      "lm_loss": 0.11559732258319855,
+      "order_loss": 51.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.209347248077393
+    },
+    {
+      "global_step": 3540,
+      "optimizer_step": 442,
+      "lr": 5.151722747865434e-07,
+      "lm_loss": 0.548110842704773,
+      "order_loss": 56.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.1731109619140625
+    },
+    {
+      "global_step": 3560,
+      "optimizer_step": 445,
+      "lr": 4.1437831084878974e-07,
+      "lm_loss": 0.2126510888338089,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.150151252746582
+    },
+    {
+      "global_step": 3580,
+      "optimizer_step": 447,
+      "lr": 3.5322453704410286e-07,
+      "lm_loss": 0.0028603943064808846,
+      "order_loss": 50.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.0653605461120605
+    },
+    {
+      "global_step": 3600,
+      "optimizer_step": 450,
+      "lr": 2.705872508804747e-07,
+      "lm_loss": 0.0031662171240895987,
+      "order_loss": 64.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.409416198730469
+    },
+    {
+      "global_step": 3620,
+      "optimizer_step": 452,
+      "lr": 2.215737794045436e-07,
+      "lm_loss": 0.052710987627506256,
+      "order_loss": 62.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.333961009979248
+    },
+    {
+      "global_step": 3640,
+      "optimizer_step": 455,
+      "lr": 1.571947526689349e-07,
+      "lm_loss": 0.008811240084469318,
+      "order_loss": 66.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 6.6650614738464355
+    },
+    {
+      "global_step": 3660,
+      "optimizer_step": 457,
+      "lr": 1.2038183319507955e-07,
+      "lm_loss": 0.005917538423091173,
+      "order_loss": 33.75,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 3.380917549133301
+    },
+    {
+      "global_step": 3680,
+      "optimizer_step": 460,
+      "lr": 7.434020362682704e-08,
+      "lm_loss": 0.032587382942438126,
+      "order_loss": 55.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.53258752822876
+    },
+    {
+      "global_step": 3700,
+      "optimizer_step": 462,
+      "lr": 4.977308833379324e-08,
+      "lm_loss": 0.0004790124949067831,
+      "order_loss": 49.25,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 4.937979221343994
+    },
+    {
+      "global_step": 3720,
+      "optimizer_step": 465,
+      "lr": 2.2125452477828047e-08,
+      "lm_loss": 0.01606699638068676,
+      "order_loss": 38.5,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 3.8598170280456543
+    },
+    {
+      "global_step": 3740,
+      "optimizer_step": 467,
+      "lr": 9.834340423678368e-09,
+      "lm_loss": 0.03486189246177673,
+      "order_loss": 50.0,
+      "ret_loss": 0.0,
+      "pretrain_loss": 0.0,
+      "total_loss": 5.034862041473389
+    }
+  ]
+}

cls_llama_baseline/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/Llama-3.2-1B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/Llama-3.2-1B-Instruct
+- lora
+- transformers
+- unsloth
+---
+
+# 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

cls_llama_baseline/adapter_config.json

@@ -0,0 +1,52 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "LlamaForCausalLM",
+    "parent_library": "transformers.models.llama.modeling_llama",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/Llama-3.2-1B-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": null,
+  "target_modules": [
+    "k_proj",
+    "v_proj",
+    "up_proj",
+    "gate_proj",
+    "o_proj",
+    "down_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
+}

cls_llama_baseline/chat_template.jinja

@@ -0,0 +1,93 @@
+{{- bos_token }}
+{%- if custom_tools is defined %}
+    {%- set tools = custom_tools %}
+{%- endif %}
+{%- if not tools_in_user_message is defined %}
+    {%- set tools_in_user_message = true %}
+{%- endif %}
+{%- if not date_string is defined %}
+    {%- if strftime_now is defined %}
+        {%- set date_string = strftime_now("%d %b %Y") %}
+    {%- else %}
+        {%- set date_string = "26 Jul 2024" %}
+    {%- endif %}
+{%- endif %}
+{%- if not tools is defined %}
+    {%- set tools = none %}
+{%- endif %}
+
+{#- This block extracts the system message, so we can slot it into the right place. #}
+{%- if messages[0]['role'] == 'system' %}
+    {%- set system_message = messages[0]['content']|trim %}
+    {%- set messages = messages[1:] %}
+{%- else %}
+    {%- set system_message = "" %}
+{%- endif %}
+
+{#- System message #}
+{{- "<|start_header_id|>system<|end_header_id|>\n\n" }}
+{%- if tools is not none %}
+    {{- "Environment: ipython\n" }}
+{%- endif %}
+{{- "Cutting Knowledge Date: December 2023\n" }}
+{{- "Today Date: " + date_string + "\n\n" }}
+{%- if tools is not none and not tools_in_user_message %}
+    {{- "You have access to the following functions. To call a function, please respond with JSON for a function call." }}
+    {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+    {{- "Do not use variables.\n\n" }}
+    {%- for t in tools %}
+        {{- t | tojson(indent=4) }}
+        {{- "\n\n" }}
+    {%- endfor %}
+{%- endif %}
+{{- system_message }}
+{{- "<|eot_id|>" }}
+
+{#- Custom tools are passed in a user message with some extra guidance #}
+{%- if tools_in_user_message and not tools is none %}
+    {#- Extract the first user message so we can plug it in here #}
+    {%- if messages | length != 0 %}
+        {%- set first_user_message = messages[0]['content']|trim %}
+        {%- set messages = messages[1:] %}
+    {%- else %}
+        {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }}
+{%- endif %}
+    {{- '<|start_header_id|>user<|end_header_id|>\n\n' -}}
+    {{- "Given the following functions, please respond with a JSON for a function call " }}
+    {{- "with its proper arguments that best answers the given prompt.\n\n" }}
+    {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+    {{- "Do not use variables.\n\n" }}
+    {%- for t in tools %}
+        {{- t | tojson(indent=4) }}
+        {{- "\n\n" }}
+    {%- endfor %}
+    {{- first_user_message + "<|eot_id|>"}}
+{%- endif %}
+
+{%- for message in messages %}
+    {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %}
+        {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' }}
+    {%- elif 'tool_calls' in message %}
+        {%- if not message.tool_calls|length == 1 %}
+            {{- raise_exception("This model only supports single tool-calls at once!") }}
+        {%- endif %}
+        {%- set tool_call = message.tool_calls[0].function %}
+        {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' -}}
+        {{- '{"name": "' + tool_call.name + '", ' }}
+        {{- '"parameters": ' }}
+        {{- tool_call.arguments | tojson }}
+        {{- "}" }}
+        {{- "<|eot_id|>" }}
+    {%- elif message.role == "tool" or message.role == "ipython" %}
+        {{- "<|start_header_id|>ipython<|end_header_id|>\n\n" }}
+        {%- if message.content is mapping or message.content is iterable %}
+            {{- message.content | tojson }}
+        {%- else %}
+            {{- message.content }}
+        {%- endif %}
+        {{- "<|eot_id|>" }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }}
+{%- endif %}

cls_llama_baseline/checkpoint-200/README.md

@@ -0,0 +1,208 @@
+---
+base_model: unsloth/Llama-3.2-1B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/Llama-3.2-1B-Instruct
+- lora
+- transformers
+- unsloth
+---
+
+# 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