Upload README.md with huggingface_hub

README.md

@@ -5,7 +5,7 @@ colorTo: red
 sdk: static
 configs:
 - config_name: default
-  data_files: "parallel_data_filtered/tld/matched_0.60_0.005_relative.jsonl"
+  data_files: parallel_data_filtered/tld/matched_0.60_0.005_relative.jsonl"
   default: true
 ---
 
@@ -38,6 +38,9 @@ In total there were three different datasets used for alignment:
  **Original Dataset:**
 - Best aligned data: `matched_0.60_0.005_relative.jsonl`
 - Total aligned document pairs: 54350
+- Romansh tokens: 42,753,157
+- German tokens: 117,237,879
+- Total tokens: 159,991,036
 - Total entries in embeddings:
   - Romansh: 208321
   - German: 106559 
@@ -56,6 +59,9 @@ In total there were three different datasets used for alignment:
   **Filtered:**
   - Best aligned data: `matched_0.60_0.005_relative.jsonl`
   - Total aligned document pairs: 53486
+  - Romansh tokens: 41,097,668
+  - German tokens: 114,789,489
+  - Total tokens: 155,887,157
 - Total entries in embeddings:
   - Romansh: 200226
   - German: 106559 
@@ -75,6 +81,9 @@ In total there were three different datasets used for alignment:
   **More Filtered:**
   - Best aligned data: `matched_0.60_0.005_relative.jsonl`
   - Total aligned document pairs: 51958
+  - Romansh tokens: 39,149,607
+  - German tokens: 110,733,850
+  - Total tokens: 149,883,457
 - Total entries in embeddings:
   - Romansh: 175195
   - German: 106559 
@@ -88,6 +97,7 @@ In total there were three different datasets used for alignment:
   - Std Dev: 0.046871048829964373
   - Threshold range: 0.50 - 0.80
 
+
 ![Threshold comparison for more filtered](threshold_comparison_tld.png)
 
 
@@ -102,8 +112,7 @@ In total there were three different datasets used for alignment:
 `penalized_sim = cos_sim - α * (abs(len_rm - len_de) / max(len_rm, len_de))`
 
 Overall best aligned parallel data: `parallel_data_filtered/tld/matched_0.60_0.005_relative.jsonl`
-
-The most filtered dataset combined with a relative length penalty (ALPHA=0.005) and a similarity threshold of 0.60 is the best aligned parallel data.
+The most filtered dataset combined with a relative length penalty (ALPHA=0.005) and a similarity threshold of 0.60 yielded the best parallel data.
 The heavy filtering removes noisy sentences before alignment, the relative penalty reduces mismatches without discarding too many valid pairs, and the 0.60 threshold balances precision and coverage. This setting consistently produced the cleanest and most reliable alignments in our comparisons.
 
 
@@ -111,7 +120,7 @@ The heavy filtering removes noisy sentences before alignment, the relative penal
 
 We tested cosine similarity thresholds (0.50–0.80) with two length-penalty settings: absolute (ALPHA=0.005) and relative (ALPHA=0.005). Across all datasets, the relative penalty consistently produced more matches, as it scales with sentence length and is less restrictive.
 
-The absolute penalty sharply reduced matches, especially above 0.55, while the relative penalty maintained higher counts but declined with increasing thresholds. In all datasets, 0.50–0.60 offered the best trade-off between quantity and precision, with the relative penalty being the more inclusive option and the absolute penalty the most conservative.
+The absolute penalty sharply reduced matches, especially above 0.60, while the relative penalty maintained higher counts but declined with increasing thresholds. In all datasets, 0.50–0.60 offered the best trade-off between quantity and precision, with the relative penalty being the more inclusive option and the absolute penalty the most conservative.
 
 
 ## Results from Dataset Analysis
@@ -120,9 +129,22 @@ The absolute penalty sharply reduced matches, especially above 0.55, while the r
 
 The domain distribution is highly imbalanced: the vast majority of URLs come from a small number of domains. Most notably, `www.rtr.ch` dominates the dataset with nearly 80,000 URLs, followed by `m.rtr.ch` and `www.gr.ch` with significantly fewer entries. All other domains appear less frequently, with fewer than 10,000 URLs each.
 
-This indicates that the Romansh dataset is heavily influenced by a few large sources, primarily public institutions and media outlets such as Radiotelevisiun Svizra Rumantscha (RTR) and cantonal websites.While this concentration can benefit consistency and linguistic quality, it may also limit the variety of content and styles represented in the dataset. This imbalance should be considered in any downstream tasks, as it could influence model performance and generalizability.
+This indicates that the Romansh dataset is heavily influenced by a few large sources, primarily public institutions and media outlets such as Radiotelevisiun Svizra Rumantscha (RTR) and cantonal websites. While this concentration can benefit consistency and linguistic quality, it may also limit the variety of content and styles represented in the dataset. This imbalance should be considered in any downstream tasks, as it could influence model performance and generalizability.
 ![Domain Distribution](domain_dist.png)
 
+### Domain Distribution of Final Dataset
+
+The final aligned dataset (51'958 document pairs) still shows the same imbalance in domain distribution. 
+As in the original corpus, `www.rtr.ch` and `m.rtr.ch` dominate the data, together accounting for 
+the majority of documents. Other domains such as `www.gr.ch`, `www.engadinerpost.ch`, and 
+`rm.wikipedia.org` contribute smaller proportions, with most domains appearing only rarely. 
+
+This again indicates that, although the filtering step likely improved the quality of parallel content, the final dataset remains highly imbalanced. 
+(See `analysis/romansh_analysis.ipynb` section _More Filtered_ for list of all domains)
+
+![Domain Distribution Final](tld_domains.png)
+
+
 ### How alignable is the dataset?
 
 After manually examining the 50 randomly sampled URLs from above, I found that none of the 50 entries had a clearly alignable German counterpart.