Merge pull request #227 from ibm-granite/testing_repo

Update model paths

notebooks/hfdemo/tinytimemixer/full_benchmarking/foundts_leaderboard/ttm-b1-512.sh

@@ -2,7 +2,7 @@ data_root_path=$1
 for cl in 512; do 
     for fl in 96 192 336 720; do
         python ../ttm_full_benchmarking.py --context_length $cl --forecast_length $fl --num_epochs 50 --num_workers 16 \
-        --hf_model_path ibm/ttm-research-r2  \
+        --hf_model_path ibm-research/ttm-research-r2  \
         --data_root_path $data_root_path \
         --fewshot 0 \
         --plot 0 \

notebooks/hfdemo/tinytimemixer/full_benchmarking/research-use-r2.sh

@@ -3,7 +3,7 @@ for cl in 512 1024 1536; do
     for fl in 96 192 336 720; do
         python ttm_full_benchmarking.py --context_length $cl --forecast_length $fl \
         --num_epochs 50 --num_workers 16 --enable_prefix_tuning 1 \
-        --hf_model_path ibm/ttm-research-r2 \
+        --hf_model_path ibm-research/ttm-research-r2 \
         --data_root_path $data_root_path \
         --save_dir results-research-use-r2/
     done;

notebooks/hfdemo/tinytimemixer/full_benchmarking/sample_notebooks/research_use/ttm-r2_freq_benchmarking_1024_96.ipynb

@@ -132,14 +132,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "# TTM models for Only Research and Academic (Non-Commercial) Use are here: https://huggingface.co/ibm/ttm-research-r2\n",
     "# Please provide the branch name properly based on context_len and forecast_len\n",
     "\n",
-    "hf_model_path = \"ibm/ttm-research-r2\"\n",
+    "hf_model_path = \"ibm-research/ttm-research-r2\"\n",
     "if context_length == 512:\n",
     "    hf_model_branch = \"main\"\n",
     "elif context_length == 1024 or context_length == 1536:\n",

notebooks/hfdemo/tinytimemixer/full_benchmarking/sample_notebooks/research_use/ttm-r2_freq_benchmarking_1536_96.ipynb

@@ -132,14 +132,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
-    "# TTM models for Only Research and Academic (Non-Commercial) Use are here: https://huggingface.co/ibm/ttm-research-r2\n",
+    "# TTM models for Only Research and Academic (Non-Commercial) Use are here: https://huggingface.co/ibm-research/ttm-research-r2\n",
     "# Please provide the branch name properly based on context_len and forecast_len\n",
     "\n",
-    "hf_model_path = \"ibm/ttm-research-r2\"\n",
+    "hf_model_path = \"ibm-research/ttm-research-r2\"\n",
     "if context_length == 512:\n",
     "    hf_model_branch = \"main\"\n",
     "elif context_length == 1024 or context_length == 1536:\n",

notebooks/hfdemo/tinytimemixer/full_benchmarking/sample_notebooks/research_use/ttm-r2_freq_benchmarking_512_96.ipynb

@@ -132,14 +132,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
-    "# TTM models for Only Research and Academic (Non-Commercial) Use are here: https://huggingface.co/ibm/ttm-research-r2\n",
+    "# TTM models for Only Research and Academic (Non-Commercial) Use are here: https://huggingface.co/ibm-research/ttm-research-r2\n",
     "# Please provide the branch name properly based on context_len and forecast_len\n",
     "\n",
-    "hf_model_path = \"ibm/ttm-research-r2\"\n",
+    "hf_model_path = \"ibm-research/ttm-research-r2\"\n",
     "if context_length == 512:\n",
     "    hf_model_branch = \"main\"\n",
     "elif context_length == 1024 or context_length == 1536:\n",

notebooks/hfdemo/tinytimemixer/full_benchmarking/sample_notebooks/ttm-r1_benchmarking_1024_96.ipynb

@@ -15,7 +15,7 @@
     "\n",
     "1. IBM Granite TTM-R1 pre-trained models can be found here: [Granite-TTM-R1 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1)\n",
     "2. IBM Granite TTM-R2 pre-trained models can be found here: [Granite-TTM-R2 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2)\n",
-    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm/ttm-research-r2)"
+    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm-research/ttm-research-r2)"
    ]
   },
   {

notebooks/hfdemo/tinytimemixer/full_benchmarking/sample_notebooks/ttm-r1_benchmarking_512_96.ipynb

@@ -15,7 +15,7 @@
     "\n",
     "1. IBM Granite TTM-R1 pre-trained models can be found here: [Granite-TTM-R1 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1)\n",
     "2. IBM Granite TTM-R2 pre-trained models can be found here: [Granite-TTM-R2 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2)\n",
-    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm/ttm-research-r2)"
+    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm-research/ttm-research-r2)"
    ]
   },
   {

notebooks/hfdemo/tinytimemixer/full_benchmarking/sample_notebooks/ttm-r2_benchmarking_1024_96.ipynb

@@ -15,7 +15,7 @@
     "\n",
     "1. IBM Granite TTM-R1 pre-trained models can be found here: [Granite-TTM-R1 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1)\n",
     "2. IBM Granite TTM-R2 pre-trained models can be found here: [Granite-TTM-R2 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2)\n",
-    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm/ttm-research-r2)"
+    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm-research/ttm-research-r2)"
    ]
   },
   {

notebooks/hfdemo/tinytimemixer/full_benchmarking/sample_notebooks/ttm-r2_benchmarking_512_96.ipynb

@@ -15,7 +15,7 @@
     "\n",
     "1. IBM Granite TTM-R1 pre-trained models can be found here: [Granite-TTM-R1 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1)\n",
     "2. IBM Granite TTM-R2 pre-trained models can be found here: [Granite-TTM-R2 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2)\n",
-    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm/ttm-research-r2)"
+    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm-research/ttm-research-r2)"
    ]
   },
   {

notebooks/hfdemo/tinytimemixer/full_benchmarking/ttm_full_benchmarking.py

@@ -4,7 +4,7 @@
 
 1. TTM-Granite-R1 pre-trained models can be found here: [TTM-R1 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1)
 2. TTM-Granite-R2 pre-trained models can be found here: [TTM-R2 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2)
-3. TTM-Research-Use pre-trained models can be found here: [TTM-Research-Use Model Card](https://huggingface.co/ibm/ttm-research-r2)
+3. TTM-Research-Use pre-trained models can be found here: [TTM-Research-Use Model Card](https://huggingface.co/ibm-research/ttm-research-r2)
 
 Every model card has a suite of TTM models. Please read the respective model cards for usage instructions.
 """

notebooks/hfdemo/ttm_getting_started.ipynb

@@ -18,7 +18,7 @@
     "\n",
     "1. IBM Granite TTM-R1 pre-trained models can be found here: [Granite-TTM-R1 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r1)\n",
     "2. IBM Granite TTM-R2 pre-trained models can be found here: [Granite-TTM-R2 Model Card](https://huggingface.co/ibm-granite/granite-timeseries-ttm-r2)\n",
-    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm/ttm-research-r2)"
+    "3. Research-use (non-commercial use only) TTM-R2 pre-trained models can be found here: [Research-Use-TTM-R2](https://huggingface.co/ibm-research/ttm-research-r2)"
    ]
   },
   {
@@ -283,7 +283,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "id": "a826c4f3-1c6c-4088-b6af-f430f45fd380",
    "metadata": {},
    "outputs": [],
@@ -295,7 +295,7 @@
     "# TTM Model path. The default model path is Granite-R2. Below, you can choose other TTM releases.\n",
     "TTM_MODEL_PATH = \"ibm-granite/granite-timeseries-ttm-r2\"\n",
     "# TTM_MODEL_PATH = \"ibm-granite/granite-timeseries-ttm-r1\"\n",
-    "# TTM_MODEL_PATH = \"ibm/ttm-research-r2\"\n",
+    "# TTM_MODEL_PATH = \"ibm-research/ttm-research-r2\"\n",
     "\n",
     "# Context length, Or Length of the history.\n",
     "# Currently supported values are: 512/1024/1536 for Granite-TTM-R2 and Research-Use-TTM-R2, and 512/1024 for Granite-TTM-R1\n",

services/finetuning/Makefile

@@ -16,7 +16,7 @@ image:
 
 clone_models:
 	git lfs install || true
-	git clone https://huggingface.co/ibm/test-tsfm mytest-tsfm || true
+	git clone https://huggingface.co/ibm-research/test-tsfm mytest-tsfm || true
 
 
 fetchdata:

services/inference/Makefile

@@ -64,7 +64,7 @@ install_dev: boilerplate
 
 clone_models:
 	git lfs install || true
-	git clone https://huggingface.co/ibm/test-tsfm mytest-tsfm || true
+	git clone https://huggingface.co/ibm-research/test-tsfm mytest-tsfm || true
 
 delete_models:
 	rm -rf mytest-tsfm || true

tests/toolkit/test_get_model.py

@@ -60,14 +60,14 @@ def test_get_model():
             assert model.config.prediction_length == fl
             assert model.config.context_length == cl
 
-    mp = "ibm/ttm-research-r2"
+    mp = "ibm-research/ttm-research-r2"
     for cl in [512, 1024, 1536]:
         for fl in [96, 192, 336, 720]:
             model = get_model(model_path=mp, context_length=cl, prediction_length=fl)
             assert model.config.prediction_length == fl
             assert model.config.context_length == cl
 
-    mp = "ibm/ttm-research-r2"
+    mp = "ibm-research/ttm-research-r2"
     for cl in range(1, 2000, 500):
         for fl in range(1, 900, 90):
             model = get_model(model_path=mp, context_length=cl, prediction_length=fl)

tsfm_public/resources/model_paths_config/ttm.yaml

@@ -88,73 +88,73 @@ ibm-granite-models:
 research-use-models:
     r2-512-96-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: main
         context_length: 512
         prediction_length: 96
     r2-512-192-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 512-192-ft-r2
         context_length: 512
         prediction_length: 192
     r2-512-336-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 512-336-ft-r2
         context_length: 512
         prediction_length: 336
     r2-512-720-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 512-720-ft-r2
         context_length: 512
         prediction_length: 720
     r2-1024-96-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1024-96-ft-r2
         context_length: 1024
         prediction_length: 96
     r2-1024-192-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1024-192-ft-r2
         context_length: 1024
         prediction_length: 192
     r2-1024-336-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1024-336-ft-r2
         context_length: 1024
         prediction_length: 336
     r2-1024-720-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1024-720-ft-r2
         context_length: 1024
         prediction_length: 720
     r2-1536-96-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1536-96-ft-r2
         context_length: 1536
         prediction_length: 96
     r2-1536-192-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1536-192-ft-r2
         context_length: 1536
         prediction_length: 192
     r2-1536-336-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1536-336-ft-r2
         context_length: 1536
         prediction_length: 336
     r2-1536-720-freq:
         release: r2
-        model_card: ibm/ttm-research-r2
+        model_card: ibm-research/ttm-research-r2
         revision: 1536-720-ft-r2
         context_length: 1536
         prediction_length: 720

tsfm_public/toolkit/get_model.py

@@ -36,7 +36,7 @@ def check_ttm_model_path(model_path):
         return 1
     elif "ibm-granite/granite-timeseries-ttm-r2" in model_path:
         return 2
-    elif "ibm/ttm-research-r2" in model_path:
+    elif "ibm-research/ttm-research-r2" in model_path:
         return 3
     else:
         return 0
@@ -63,10 +63,10 @@ def get_model(
             model name to use. Allowed values: ttm
         context_length (int):
             Input Context length. For ibm-granite/granite-timeseries-ttm-r1, we allow 512 and 1024.
-            For ibm-granite/granite-timeseries-ttm-r2 and  ibm/ttm-research-r2, we allow 512, 1024 and 1536
+            For ibm-granite/granite-timeseries-ttm-r2 and  ibm-research/ttm-research-r2, we allow 512, 1024 and 1536
         prediction_length (int):
             Forecast length to predict. For ibm-granite/granite-timeseries-ttm-r1, we can forecast upto 96.
-            For ibm-granite/granite-timeseries-ttm-r2 and  ibm/ttm-research-r2, we can forecast upto 720.
+            For ibm-granite/granite-timeseries-ttm-r2 and  ibm-research/ttm-research-r2, we can forecast upto 720.
             Model is trained for fixed forecast lengths (96,192,336,720) and this model add required `prediction_filter_length` to the model instance for required pruning.
             For Ex. if we need to forecast 150 timepoints given last 512 timepoints using model_path = ibm-granite/granite-timeseries-ttm-r2, then get_model will select the
             model from 512_192_r2 branch and applies prediction_filter_length = 150 to prune the forecasts from 192 to 150. prediction_filter_length also applies loss